*** Complete AMPL fix and change log *** Below is a summary of AMPL bug fixes and changes from 30 Nov. 1992. Note that AMPL versions are strings of the form yyyymmdd, giving the date (year = yyyy, month = mm, date = dd) of the last change affecting the version. You can determine the version of your "ampl" program by issuing the AMPL command option version; or by invoking ampl with the -vv option. To see the version without executing ampl further, use the invocation ampl -vvq An up-to-date summary of AMPL changes and recent bug fixes is available by electronic mail: send netlib@netlib.bell-labs.com the E-mail message send changes from ampl To get the current version of this file, ask netlib to send fixlog from ampl Here is a summary of bug repairs and changes. Each block of changes is preceded by the version (yyyymmdd) in which the changes first appeared. 19921130: Fix bug with option omit_zero_rows: numeric set members were sometimes treated as zero. Example: option omit_zero_rows; display{i in -3..3} i; 19921210: Fix bug with print and printf illustrated by print{1..3}: 1; which only printed 1 once. Fix bug in detecting common expressions in function arguments. Example: function nfcall(); print{i in 1..3} ncall(); used to print "1 2 3"; now prints "1 1 1" (where ncall returns its invocation count -- it's one of the examples in the sample funcadd.c). 19921216: Fix horrid bug in (and following), e.g., assignment to a new param of computed values involving a variable other than the first-declared variable. 19921218: Fix possible memory fault with objective.invalid_suffix. Fix possible confusion that might have resulted if changed data caused constraints, objectives, or variables to be regenerated after some components were dropped or fixed. For now [until 19930813], partial drops and fixes are forgotten when an entity is regenerated. Complete drops/fixes ("drop foo;" or "fix goo;") are still honored. Add initial version of "show" command: "show;" or "show >filename;" lists all model entities. "show name;" or "show name >filename;" shows name's declaration if it has one, or else lists model entities of the kind indicated by the first letters of name: ch... ==> checks c... ==> constraints f... ==> functions o... ==> objectives p... ==> parameters s... ==> sets v... ==> variables 19921221: Fix bug in linearizing "easy" piecewise-linear terms (convex terms in minimizing objectives, concave in maximizing, etc.) with negative breakpoints. 19921231: Fix bug in ampl -omfoo model steel.mod; data steel.dat; solve; write; # "solve;" had destroyed command-line outopt setting Fix bugs in reset data and reading data for parameters that have a default. Example: set A := 1..3; var x{A}; param b{A} default Uniform01(); minimize zot: sum{i in A} (x[i] - b[i])^2; solve; data; param b 2 27; # no complain about b already having # a value computed from its default reset data b; solve; # the model was not instantiated anew Adjust "update data" command to retain previously computed default values unless explicitly assigned in a subsequent data section. This mainly affects random default values. Fix bugs in the error messages for parameters that have previously computed default values and (absent "update data" permission) are given new values in a data section. 19930103: Fix bugs with params that have a default that gets evaluated before a data section provides values for some other components, e.g.: param p{1..10} default Uniform01(); display{i in 9..10} p[i]; data;param p 1 3 2 37; display p; # wrong values for p reset data p; display p; data; param p 3 23 5 42; display p; # memory fault 19930111: Fix obscure bugs: (1) If let or a data section changed a variable that had been fixed by presolve, subsequent solve commands (without some other intervening event to provoke running presolve again) did not restore the variable. Example: var x; var y; minimize obj: (x - y)^2; c1: x + y >= 7; c2: x = 0; let x := 9; let y := 11; # x correctly set to 0 solve; display x,y; let x := 9; let y := 11; solve; display x, y; # x was left at 9 (2) Constraint.body (etc.) values were not correctly computed after "let" or a data section changed a variable that had been fixed by presolve and appeared nonlinearly in the constraint, and similarly for objective values. Adjust logic for deciding if "solve" needs to rewrite the .nl file (e.g. after a "write" command). Previously, changes to variables by "let" examined $reset_initial_guesses at the time of the let command; now ampl notes that there are changes, and examines the state of $reset_initial_guesses and $dual_initial_guesses when it is time to consider rewriting the .nl file. Fix bug with "show" revealed in var x{1..3}; minimize zot: sum{i in 1..3} (x[i]-i)^2; solve; display x; show x; display x; # wrong values 19930112: Fix infinite loop that could result from typing "end" (rather than "quit") after certain error messages. Keep end-of-file in data mode from causing instantiation of the current problem if commands have been seen. 19930114: Fix bug with -on (or option outopt n...). 19930119: Fix bug revealed in var x{1..2}; minimize obj: sum{i in 1..2} (x[i] - i)^2; solve; show x; let {i in 1..2} x[i] := 10; # memory fault Fix bug revealed in var x{1..4}; let {i in 1..4} x[i] := i; # rhs = unadorned dummy variable display x; # wrong values Tweak some details of printing by "show": no more ".if." or "recpd."; show recursive parameter defs with := rather than "default" (when the declaration specifies :=). Adjust rule for making "if" symbolic or numeric. This should be invisible, except that var x{i in 1..4} := if i = 1 then i else x[i-1]+2; now elicits an error message about "variable in := expression". 19930126: Adjust solve command to notice changes in $reset_initial_guesses and $dual_initial_guesses since the last write command. Fix bugs in update data: 1. Some derived sets were not cleared properly. 2. Check statements were not run again after data updates. This could cause some derived sets not to be recomputed. 19930213: Coerce if (expression involving variables) to if (expression != 0) rather than complaining about variables in a logical expression. Supply missing subscript in error message for set A; param p{A}; data; param p := a 1 a 2; 19930208: Add new option relax_integrality: option relax_integrality 1; causes "integer" and "binary" attributes of variables to be ignored (in solve and write commands). 19930209: Adjust interlocks so changing option var_bounds after a write command will cause a subsequent solve command to write temporary files (to pass the alternate variable bounds to the solver). 19930303: Fix glitch in constraint.lb, .ub, .lbs, .ubs: constraint.lb and .ub do not require presolve, .lbs and .ubs do. Variable.lb3, .ub3 withdrawn. Add logic and option constraint_drop_tol (default 0) to deal with subtle presolve bug apparently caused by roundoff error: with $presolve > 1 and $var_bounds == 1 (the defaults), constraint bounds were very occasionally relaxed due to bounds only conveyed for $var_bounds > 1; this could increase the size of the feasible region, possibly making the problem unbounded. (Only known example: test problem MAROS from netlib's lp/data.) The fix involves keeping two sets of constraint bounds and switching between them based on $var_bounds. The constraint bounds for $var_bounds == 1 are only relaxed if roundoff poses no danger or the deduced bounds on the constraint body are sharper than the declared bounds by at least $constraint_drop_tol. (The default $constraint_drop_tol value 0 causes both sets of constraint bounds to be the same and gives the same behavior as before this change.) New constraint dot values: constraint.lbs1, .ubs1, lbs2, .ubs2 = versions of .lbs, .ubs corresponding to $var_bounds <= 1 or > 1, respectively. Constraint.lbs, .ubs still reflect the bounds sent to the solver. Fix bug (memory fault) in use of constraint[subscript].suffix before a solve or solution command. Delete warning about possibly incorrect dual values inferred for constraints eliminated by presolve: the inferred dual values are now believed to be correct in all cases. 19930311: Fix some bugs in handling defined variables after "let" changes the problem's size. 19930315: Fix bug (giving error msg "invalid refct 0 in opgen") revealed in param M{1..2}; param t := M[1] / (M[1] + M[2]); param n := M[1] + M[2]; param p{1..n} := Uniform01(); var x{i in 1..2}; minimize zot: (sum{i in 1..2} (sum{j in 1..n} p[j])* (x[i] - M[i])^2)/(M[1] + M[2]); data; param M := 1 1 2 2; print zot; print t; let M[1] := 3; print zot; Adjust presolve to use directed roundings on (some) IEEE-arithmetic machines. This sometimes leads to fewer surprises, such as an unbounded objective under default conditions with netlib's "maros from lp/data". 19930319: Fix bug after the message "Ignoring solve command because presolve finds no feasible solution possible." Neither reset nor drop cleared the "infeasible" flag. 19930331: Fix bug in printf's handling of %e, which rounded to one too few decimal places. 19930406: Fix bug (mix-up of convex versus concave) in linearizing piecewise-linear constraints. Convex <= constraints had integer variables added unnecessarily, and convex >= constraints did not have them added. Fix a bug that currently only matters to osl when a problem contains some piecewise-linear terms that require adding integer variables and others that do not require this. 19930409: Fix horrid bug in handling slices after "let" or "reset data" changes the slice's base set. Some slices were not recomputed. Example: set A dimen 2; set I := setof{(i,j) in A} i; set B{i in I} := {(i,j) in A}; data; set A := (a,b) (a,c) (b,c); display B; reset data; data; set A := (a,y) (w,z); display B; The second display printed set B[a] := b c; set B[w] := ; # empty rather than the correct set B[x] := y; set B[w] := z; Fix a bug that sometimes appeared after reset data when e.g. variables or constraints are indexed over a setof or over a union or subscripted set whose size was a power of 2. 19930421: Fix bug in "let" assigning 0 to a param after an error message about the param not having a value. 19930422: Fix rare memory allocation bug that might be encountered with printf format %.24e (on systems with 32-bit ints; for 16-bit ints, %.16e could elicit the bug). [For the MS-DOS Student Edition, AMPL.EXE has 32-bit ints, and AMPLC.EXE has 16-bit ints.] 19930506: New command-line option -v2 merges stderr with stdout both for ampl and for subprocesses (whereas -ve just merges them for ampl). Allow negative precision with %f, with formatting as for the print command with $print_round < 0. Fix botch with printf %wg (w = integral width): integers had as many extra spaces as trailing zeros. 19930514: Added an fflush(stderr) that affects placement of error messages in invocations like ampl foo >goo 2>&1 and ampl -v2 2>foo (which should result in an empty foo). Fix bugs revealed in c1: to_come = 1; param p; var x coef c1 p := 1/3; data; param p := 3; let p := 2; print c1.body; print p; print x; #garbage and model prod.mod data prod.dat solve; reset data; data prod.mod # wrong file reset data; data prod.dat solve; # memory fault Arrange for errors that stop genmod to terminate current file inclusions (as other errors do); close the files involved. Close current input file right after EOF, rather then just before opening a new (command-line) input file. 19930515: Adjust logic so "let" can reference current variable values without instantiating an updated problem instance (which could elicit complaints about missing values -- being supplied by "let"). Avoid a memory fault when a command references an unavailable subscripted set after the first error message about its absence. For entities with several subscripts, enforce the rule that literals must be quoted in the model. The first printing of thet AMPL book (middle of p. 252) erroneously shows some unquoted literals, which the AMPL translator should not have accepted. 19930521: Fix bugs with "objective" and "drop" that caused these commands to be ignored when applied to simple objectives or constraints involving to_come, net_in, or net_out. Fix bug in option Cautions, which did not suppress Cautions about multiple "to" and "from" clauses issued during genmod. Modify printing of objectives (by "show" and -M) so minimize cars; prints as on the previous line, rather than (the equivalent) minimize cars: 0 + to_come; 19930609: Fix memory fault revealed in the sequence model diet.mod data diet.dat reset data; data diet.mod # elicits many error messages reset data; # memory fault 19930615: Fix bug in writing .fix file: if the .unv file was not requested and there were unused variables, they went into the .fix file. Fix bug in reporting values of unused variables: they were reported as having their initial guess, which might be infeasible (i.e., might not satisfy their bounds). Now they're adjusted to be in bounds. Use stronger deduced variable bounds rather than declared variable bounds in deciding whether an integer variable is binary. Fix error message that complains about duplicate entries in subscripted sets read in a data section: the subscripts were wrong and could cause a memory fault. 19930616: Fix bug in printing the subscript of an indexed set in certain error messages. Example: set A; set B{A}; set C := union{i in A} B[i]; param p{C}; data; set A := a b; set B[a] := 1 2; # no set B[b]; display C; # Bus error while complaining about B[b] 19930630: Fix bug introduced 8 Feb. 1993 (with relax_integrality, version 19930208): option presolve 0 ignored integrality. Adjust .nl files to convey (in their first 10 lines) counts of nonlinear integer variables and of nonlinear variables used in both constraints and objectives (nlvb). When nlvo > nlvc, arrange for the first nlvb variables to be the nonlinear variables common to both constraints and objectives (and the next nlvc - nlvb to be just in constraints). Adjust tables in "Hooking Your Solver to AMPL" to describe the new ordering of nonlinear variables; Postscript for this revised report is available by E-mail: ask netlib@netlib.bell-labs.com to send 93-10 from research/nam 19930702: Fix bug (e.g., memory fault) sometimes revealed in solve; display foo; where foo is a simple or indexed variable, none of whose components are used in the model. 19930711: Fix bug with defined variables: they were not recomputed after "let" should have caused them to change. Fix bug in recompiling "if" expressions whose "then" or "else" clauses involve variables. Recompilation is necessary after some "let" and "reset data" commands. 19930711: Fix bug in restoring variable values when regenerating an instance (e.g. after "let") whose model has defined variables declared before some non-defined variables: the subsequent non-defined variables got some wrong (seemingly random) values. 19930712: Fix bug with -L (option linelim 1): nonlinear uses of a linear defined variable did not cause the right-hand side variables to be recorded as nonlinear variables. Example: var x; var y = x + 4; var z = y^2 + 2*y + 1; minimize zot: z^2; # x not recorded as nonlinear # bug caused gradient == 0 19930805: New option presolve_inteps (default 1e-6) gives tolerance for rounding updated bounds on integer variables to integer values during presolve: if x.dlb and x.dub denote the new deduced lower and upper bounds on x, then for $presolve_inteps < 1, x.dlb := ceil(x.dlb - $presolve_inteps) and x.dub := floor(x.dub + $presolve_inteps). For $presolve_inteps >= 1, x.dlb := floor(x.dlb) and x.dub := ceil(x.dub). Adjust presolve complaints about impossible deduced bounds to take $presolve_eps into account. 19930813: Change ">=" to ">" in error message "lower bound = ... > upper bound = ...". Fix nasty bug with dropping only some components of an indexed constraint: unless only the final constraints were dropped, the constraints got the wrong left- and right-hand side bounds. Arrange for $relax_integrality to apply to declared bounds (which formerly got rounded to integer, even if option relax_integrality 1; was specified). Apply $presolve_inteps to the rounding of bounds declared with variables (when $relax_integrality is 0). Fix memory fault (or worse) occasionally seen after drop or restore. Omit interlock that gave the message "x.rc cannot be used until after a solve or solution command", which (confusingly) could be circumvented by displaying x.rc first. Stop iterated "let" at first invalid left-hand side subscript. Adjust bounds on integer variables to tightest available bounds (thus always passing bounds of 0 and 1 for variables that the solver is told are binary variables). Retain dropped status of constraints, fixed status of variables, and current values of scalar variables and dual values of constraints when the problem changes. (Current values of subscripted variables were already retained. Dual values were previously discarded.) 19930901: Fix memory fault in set S := {1..5} cross {3..9}; display {(i,j) in S: i=j}; display {i in 1..5}: {(i,j) in S}; # memory fault Fix optimization bug: constraints or objectives involving a context of the form {i in A, j in B: condition(i,j)} and a linear expression of the form ("constant" expression involving i but not j) * (expression involving variables) did not have ("constant" expression involving i but not j) re-evaluated when i changed but condition(i,j) was false for the first j in B. Example: param p{i in 1..2, j in 1..2} := 10*i + j; var x{i in 1..2} >= 0; minimize zot: sum{i in 1..2, k in 1..2: k > 1} (sum{j in 1..2} p[i,j]) * sum{L in 1..2} x[L]; Also fixed: memory fault when the above example is fed to "ampl -O". 19930914: Fix bugs in handling "if" expressions yielding symbolic (character) values used as arguments to functions involving variables (i.e., functions that the solver must evaluate). Example: function foo; var x; minimize zot: foo(x, if x > 3 then 'abc' else 'def'); Fix bug in handling defined variables whose right-hand sides involve (linear occurrences of) piecewise-linear terms when option pl_linearize has its default value 1. Fix storage-overwriting bugs that sometimes arose when an entity was indexed over a set whose cardinality changed from 1 to some larger value (as a result of "let" or "reset data"), or when the set over which a defined variable is indexed similarly got (sufficiently) larger. Add automatic differentiation facilities for computing reduced costs of variables in nonlinear problems (and dual values of definitional constraints and constraints eliminated by presolve). This eliminates the old message "Ignoring nonlinearities in computing dual values for constraints eliminated by presolve." New dot notation variable.dual applies to defined variables and gives the dual value for the defining equality constraint; this gives the partial derivative of the Lagrangian function (objective minus sum of duals times constraints) with respect to the defined variable. For other variables, variable.dual = 0. 19930922: Fix bus errors and erroneous "bad subscript" messages revealed by param n integer >= 0; set S := 1..n; set T; param p {T,S} >= 0; check {i in T}: sum {j in S} j * p[i,j] <= 2; var x{S,T} >= 0; minimize zot: sum{i in S, j in T} x[i,j]; data; param n := 0; set T := a b c; let n := n + 1; let p['a',n] := 1/n; let{j in T: j != 'a'} p[j,n] := 2/n; solve; [The "check" first executes at the first "let", then again at the "solve"; there was a bus error in preparing this second execution of the "check". There was also a bus error associated with the references to n in commands after n had been incremented. And there was a bug in handling sets like S that go from being empty (n = 0) to having members.] 19930928: Fix bug in handling a variable indexed over a computed set: if the set changed and had to be recomputed before the variable was re-instantiated, saved values could be wrong and (if the set got sufficiently bigger) a memory fault could result. Example: param n default 2; set S := n..n+2; var x{i in S}; let{i in S} x[i] := 10*i; display x; let n := n + 1; display S; # forces S, but not x, to be recomputed display x; # wrong values -- subscripts not adjusted Fix bug in subscripts shown in error messages for failed check clauses in param declarations (for params with values given in a data section). Apply $solution_precision and $solution_round to dual as well as primal solution values. Fix printf buglet: printf "%04.2d\n",3 printed 0003 rather than 03. Fix a bug in solving a problem, increasing the index set of a variable beyond the next power of 2, and solving a modified problem. Fix bugs in combining "option presolve 0" with "fix" and with "fix" before "let", "solve", or "solution" changes variable values. Example: var x{1..2} >= 0; minimize zot: sum{i in 1..2} i*x[i]; s.t. convex: sum{i in 1..2} x[i] = 1; let x[1] := .3; fix x[1]; # was ignored if previous "let" was omitted option presolve 0; solve; # Memory fault; --------------- FLOW OF CONTROL --------------- Several new commands permit conditional execution of and looping over lists of AMPL commands: if lexpr then cmd if lexpr then cmd else cmd # else matches nearest available if for opt_name indexing cmd # dummies from indexing may appear in cmd repeat opt_name opt_while { cmds } opt_while ; break opt_name ; continue opt_name ; cmd is either a single command (ending with ;) or { cmds } . cmds is a sequence of 0 or more commands . lexpr is a logical expression. opt_name is an optional loop name (which must be an unbound before the syntactic start of the loop), which goes out of scope after syntactic end of the loop. opt_while clauses are optional. If not null, opt_while has the form while lexpr or until lexpr If the optional loop name is not specified, break and continue apply to the immediately enclosing loop; otherwise they apply to the named loop; break terminates the loop, and continue causes its next iteration to begin (if permitted by the optional initial and final opt_while clauses of a repeat loop, or by the indexing of a for loop). Loops and if-then-else structures are treasured up until syntactically complete. Because else clauses are optional, AMPL must look ahead one token to check for their presence. At the outermost level, one must thus issue a null command (just a semicolon) or some other command or declaration to execute an outermost else-less "if ... then stmt". (In this regard, end-of-file implies an implicit null statement.) New options cmdprompt1 and cmdprompt2 (called prompt3 and prompt4 until 19930423) control prompting within the new flow-of-control commands. 19930930: Fix unlikely bug in constraints omitted by an "indexing" of the form {if false_expression}: the bug only appeared if certain virgin memory was nonzero in the right way. Fix bug in handling variables in flow-of-control tests before any solve, write, solution, or printing command. Example: var x; if x > 0 then print 'Yikes!'; Fix long-standing memory-overwrite bug in "show" (when listing the names of all entities of a class, such as all variables or all constraints) that only bit occasionally. 19931005: Fix bug in recovering dual values for constraints eliminated by presolve when $solution_round or $solution_precision caused the solution to be rounded: the rounding occurred before the dual values were computed, sometimes leading to wrong decisions about which constraints were active. New options $abs_boundtol, $rel_boundtol, and $show_boundtol are meant to help deduce correct dual values for constraints eliminated by presolve when the solver uses an interior-point algorithm and returns a solution with no bounds strictly holding. All three new options have default value 0, which gives the previous behavior. Suppose for some variable x that the solver sees the bounds lb <= x <= ub. The lower-bound constraint lb <= x is considered active (during reconstruction of dual values) if x <= lb or (x - lb < ub - x and x - lb <= max($abs_boundtol, |lb|*$rel_boundtol)), and similarly for the simple upper-bound constraint (x <= ub). Thus negative values of $abs_boundtol and $rel_boundtol behave like 0. The condition x - lb < ub - x ensures that x is closer to lb than half-way between lb and ub, ensuring that AMPL picks the more appropriate bound no matter how large $abs_boundtol and $rel_boundtol are. New option $show_boundtol works similarly to $show_stats, except that it delivers its messages when it is on (nonzero) and another dual-value computation occurs or (like $show_stats) when it is set to 1. It reports changes to $abs_boundtol and $rel_boundtol that would change the outcome of the dual computation, and is silent if the values of $abs_boundtol and $rel_boundtol do not matter. [$show_boundtol was called $show_boundstats until 20 Dec. 1993.] Have option redirections affect printing for "option show_stats 1;" and "option show_boundtol 1;". Fix bug in computing constraint.ldual and constraint.udual for constraints after the first one: the decision whether to the constraint is binding was in error and could memory fault. Fix bug in handling constraints and objectives declared with syntax errors: references to them were botched. Fix memory faults with various variable.suffix and constraint.suffix notations before the first solve or other command that caused the model to be instantiated fully enough. 19931006: Fix bug in handling ord(dummy), next(dummy), prev(dummy), next(dummy,n), and prev(dummy,n) after "reset data;". Example: set S ordered; set T := {i in S, j in S: ord(i) < ord(j)}; data; set S := A B D; display T; reset data; data; set S := A B C D; display T; # complained that B was not in S. Forms that explicitly specify the set involved were handled correctly, e.g.: set T := {i in S, j in S: ord(i,S) < ord(j,S)}; 19931012: Arrange for errors in "let" commands to terminate all "for" and "repeat" loops. Fix bugs in use of "first", "last", "card" in "for" and "repeat" loops. Fix bugs in "ordered by" for sets given values in a data section. Example: set A ordered; set B ordered by A; data; set A := a b c; set B := b c a; display B; # memory fault Fix bug in "update data S" where S is a set. Example: set A; set B default A; data; set A := a b c; display B; update data B; data; set B := c a; # erroneous complaints of duplicate members display B; Under option relax_integrality, infer a lower bound of 0 and upper bound of 1 for binary variables. Fix bug in handling sequence "if ... then {...} if ...": the second if should cause the first to be complete (have a null else part). Example: if 1 > 2 then print 3; if 4 < 5 then print 6; else print 7; should print 6; with the bug, it did nothing. Fix bug that caused default expressions not to be evaluated in situations with incomplete "if" commands. Example: param p default atan(1); param p; if 1 < 2 then display 3; let q := p + 1; # memory fault Fix bug in handling dropped constraints after "let" or "update data" causes changes to some constraints. The dropped constraints came back into the problem. 19931018: Fix bug in references to variable.lb or variable.ub for hitherto unreferenced variables (after, e.g., "solve" or "display" has caused some entities to be instantiated) that do not appear in any constraints or objectives. Fix memory overwrite bug sometimes seen in displays involving more than 4000 characters. 19931022: Fix an obscure bug that arose in a complicated "display" before "solve" indirectly involving "prev" or "next". A simple example seems hard to create. 19931029: Arrange for expressions involving primal and dual variable values to be recomputed when those values change. Example: set S := 1..6; var x; for{i in 2..4} { let x := i; display {j in S: j > x.val}; } formerly displayed the same set thrice; now each is distinct. Treat variable as variable.val, constraint as constraint.dual in indexing expressions for "fix", "unfix", "drop", "restore". Fix bug (memory fault) introduced 6 Oct. 1993 in handing some next() and prev() references. Fix bug in handling sets involving dummy variables instantiated by for{...}. Example: for{i in 2..8 by 3} print{j in i..i+2} j; formerly printed "2 3 4" thrice. New syntax for fix and unfix commands: an optional := expr may appear before the terminating semicolon, in which case the expression is assigned to the variable being fixed or unfixed (as though assigned by "let"). New option ampl_include gives a white-space separated list of directories in which to search for files in "include", "model", and "data" commands. In this list, a single period stands for the current directory. The default, '.' (a single period) gives the same behavior as heretofore. References to absolute file names (starting with "/" or, for MSDOS, one of "/", "\", or "x:", where x is any printing character) are not affected by $ampl_include . 19931113: Add "exit" as synonym for "quit". Recognize file names that start with "./" as file names relative to the current directory (and ignore $ampl_include for such file names). Arrange for all (?) expressions involving dot notation to be recomputed when the "dot value" changes. This involves adding "system" parameters _Solution_Count and _Initial_Guess_Count that, for debugging, may be referenced as params (but which may become invisible later). New option presolve_intepsmax (default 1e-5). The message "Setting $presolve_inteps >= nnn might help" is suppressed if nnn > $presolve_intepsmax. New option presolve_warnings (default 5) limits the number of warning messages printed during presolve; subsequent warning messages are suppressed, and the number of suppressed warnings (if positive) is reported at the end of presolve. When $presolve_warnings < |$eexit| (as is true by default), a subsequent "Ignoring solve command because presolve finds no feasible solution possible" may now appear even when presolve finds at least |$eexit| causes for infeasibility. 19931123: New option log_file (default ''). If $log_file is not '', then all lines read from stdin or written to stdout or stderr are copied to file $log_file. Fix bug (perhaps introduced 13 Oct. 1993) in checking changes to subscripted sets that are ordered by or within another set, and which have a different number of subscripts than their dimension. 19931201: Fix memory fault in param p symbolic; let p := 3; # core dump Correct offsets in error messages in PC versions (which with some compilers were corrupted by MS-DOS's \r\n idiocy). 19931203: Fix bug with "let" commands inside "for" or "repeat" loops, when "let" assigns the same set to two or more other sets. Adjust error messages to mention the command being executed (and, if the command came from a file, to mention the file, line number, and offset) when an error occurs in a flow-of-control context, such as a loop. New option bad_subscripts: ampl now discards invalid subscripts (read in a data section or assigned by "let"), and the accompanying error message now shows at most $bad_subscripts (default 3) invalid subscripts per entity (when there is more than one bad subscript). 19931205: Tweak error message for aborted solve commands on Unix systems. 19931206: Fix memory fault in referencing (e.g.) an indexed param after "update data" if the param had no value before "update data". Example: set A; param p{A}; update data; data; param :A: p := a 1 b 2; display p; # memory fault 19931217: Fix bug in handling card, first, last in certain complicated situations. Example: set A; set B{A}; param p{i in A} := sum {j in B[i]} card({i} union {j}); # memory fault when p was instantiated Fix some memory leaks in solving sequences of problems. Fix memory fault that resulted when $log_file was initially set. 19931220: Fix glitch in changes of 13 Nov. 1993 (version 19931113) that caused the second solve in the following (pathological) example to have optimal value 3 rather than 6: var x >= 0; param a := x.val; maximize obj: x; subject to con: x <= a + 3; solve; solve; Fix an obscure memory fault. Change $show_boundstats to $show_boundtol. Retain dual values deduced for constaints eliminated by presolve. They were hitherto lost when a new problem was instantiated. 19931230: Fix memory fault in "reset data;" introduced in version 19931113. 19940103: Fix glitch introduced in 19931203 in the file name reported by error messages: in the invocation ampl foo - error messages for commands issued from stdin claimed to be in file foo. 19940113: Fix bug that prevented drop and restore from recognizing objectives. 19940119: Fix erroneous printing of a file name in error messages for commands typed on stdin after "reset". Fix occasional omission of file name from error messages (under complicated circumstances). Adjust a presolve error message so var x; fix x := 2; s.t. zot: x^2 <= 3; solve; gets the error message all variables eliminated, but upper bound = -1 < 0 rather than no variables, but lower bound = -Infinity, upper = -1 which is now reserved for constraints like s.t. zot: 0 <= -1; 19940128: Fix bug (possible loop or worse) in handling decimal numbers with an outrageously negative exponent field, such as 1.8826e-512. Adjust hashing to work better on {1..25,1..25,1..25,1..25}. Add column headings to output for -t and -T, and added incremental memory column to -T. Fix memory fault resulting from an attempt to display something about a hitherto unused variable after "solve". Example: var x; var y; minimize zot: (x-2)^2; solve; display y.lb; Fix rarely seen bug in the interaction of option show_stats and redirections on printing commands: if a printing command forced presolve to run (with "option show_stats 1" in effect), the redirection got lost and the printing went to stdout. 19940203: Fix bug that appeared under circumstances for which giving a short example seems hard: after a "solve" and printing variables that only appear in dropped constraints, printing components of some variables (before a display that printed all their values) sometimes showed the wrong values for them. Under similar circumstances, a memory fault (or worse) was possible with nonlinear defined variables. Fix possible memory fault in card expressions that involve a dummy index for a collection of constraints, objectives, checks, column generation syntax, or iterated function arguments. Example: set A; param p{A}; set B; var x{A,B}; s.t. zot{i in A}: sum {j in B} x[i,j] = card{j in A: p[j] = i}; # zot suffered a memory fault in "compile". 19940207: Fix an obscure memory fault illustrated by set P := 1..2; param d { P,P }; set PP := {P,P}; s.t. symmetry {p1 in P, p2 in P}: d[p1,p2] == d[p2,p1]; # Silly constraint: should be a check. var phi { P } >= 0; minimize pot: phi[1]; for{i in 1..2} { for {(p1,p2) in PP: p1 <= p2} { let d[p1,p2]:= Uniform01(); let d[p2,p1] := d[p1,p2]; } display symmetry.slack; # fault when i = 2 } 19940208: Fix bug introduced in 19940128 in handling subscripted sets of arity >= 2 and more than 16 elements that are only used for membership tests: the membership tests could come out wrong. Example: param n integer > 0; set A{1..n} dimen 2; set B{1..n} dimen 2; set C{i in 1..n} := A[i] union B[i]; set D{1..n}; set E{i in 1..n} := {j in D[i]: (j,j) in C[i]}; #silly let n := 1; let A[1] := setof{i in 1..50, j in i-1..i+1} (i,j); let B[1] := setof{i in 45..130, j in i-2..i+2} (i,j); let D[1] := 40..55; display E; # erroneously empty; should be 40..55 19940210: Fix fault in use of next and prev in some contexts. Example: set s circular := 1..7; param b {s}; let {i in s} b[i] := i; let {i in s} b[i] := b[next(i)]; # segmentation fault Fix remotely possible bug with option log_file and "reset;". 19940217: Fix obscure bug with output style 'm' (option outopt m... or command-line option -om...): initial guesses (in the form expected by MINOS: "FR INITIAL" lines in the BOUNDS section) were omitted from MPS files for linear problems. Fix "solveout bug" message that could arise under complicated conditions with -L ("option linelim 1;"). Fix memory fault possible with -L and variables defined by a linear expression plus nonzero constant. Fix bug, under -L, with nonlinear defined variables that have linear terms and are only used linearly: if there were no nonlinearly used nonlinear defined variables with linear terms, an erroneous .nl file resulted (and ampl could fault). Fix possible error associated with options OPTIONS_IN and OPTIONS_INOUT in the file name shown in error messages for commands. Fix bug in interaction of option outopt (or the write command) and certain display commands that could cause extra work (an an extra display of presolve statistics if "option show_stats 1" is in effect) when the session ended with end-of-file (rather than a quit command). 19940303: Continue execution when either $OPTIONS_IN or $OPTIONS_INOUT is unreadable at the start of execution. Allow any option value that does not need to be quoted in a data section to be unquoted in option commands. Option values have always been printed without quotes when quotes can be elided in a data section, which made it impossible for $OPTIONS_INOUT to restore a value like a.b (or just ".", the default value for $ampl_include). Side effect: numeric option values are no longer rounded (to the shortest decimal string rounding to their numerical value rounded to the machine's arithmetic). For example, previously option foo 00123, goo '00123', zoo 1.234567890123456789; option foo, goo, zoo; printed option foo 123; option goo 00123; option zoo 1.2345678901234567; and now it prints option foo 00123; option goo 00123; option zoo 1.234567890123456789; 19940317: Fix possible memory fault under "ampl -M" with cross products. Example: set ORIG; # origins set DEST; # destinations set PROD; # products set orig {PROD} within ORIG; set dest {PROD} within DEST; set links {p in PROD} := orig[p] cross dest[p]; Side effect: A cross B now prints as A cross B rather than {A,B}. Turn {A cross B} into {A,B}. Do not let "let" assign set values of the wrong arity. Fix bugs with "reset data foo;", where foo is a variable or constraint (a rarely used feature: updating .init and .init0 values). Avoid generating unused variables in response to printing commands (that leave the variables unused). Allow "let S := {};" for sets S of arbitrary arity. 19940329: Fix bug in handling if expressions involving a symbolic then and numeric else part, or vice versa: coerce the numeric part to symbolic, rather than vice versa. 19930401: Fix storage allocation bug revealed by complicated interaction of let with a variable indexed by a growing set. Example (from Leslie Hall, giving an erroneous error message about an invalid subscript): modified data for cut.run from Chvatal exercise 13.2b: param roll_width := 181 ; param: WIDTHS: orders := 21.625 90 20.5 51 20 45 17.25 11; 19930423: Fix error messages for "can't multiply ... by NaN" and "can't multiply ... by Infinity" to get subscripts in ... right. Fix error message for bad subscripts in column-generation notations (including arc declarations) to get subscripts right. Have the reset command close pipe functions. Fix bug in handling option commands after "solve" in loops: the name and value of the option were sometimes overwritten. Fix buglet that could force values returned by a solver subsequently to be forced in bounds when a subsequent "let" or "update data" causes more variables to be generated. This has a minor effect on example looping/multi.run (change by one in number of iterations in one of the solves). Allow any UTF character beyond the 7-bit ASCII characters to appear in names. Fix nasty bug in generating command prompts after "reset;": if command prompts for named loops were generated before the reset, such prompts generated after the reset could scribble on memory now being used for something else. Change prompt3 and prompt4 to cmdprompt1 and cmdprompt2. New options dataprompt1 and dataprompt2 are analogous to prompt1 and prompt2, but for data mode; defaults 'ampl data:' and 'ampl data?'. Catch SIGINT ("break" or "del" key). When received, terminate reading all files except stdin, and abort compound commands. Stop if a second SIGINT arrives before a successful read on stdin. 19940429: Abort compound commands when solve returns a nonzero status (e.g., if the solver was stopped by a SIGINT). 19940506: Test whether variables fixed by the "fix" command lie within $presolve_eps of their declared ranges. 19940512: Fix bug that, under rare conditions hard to describe, caused defined variables sometimes to have the wrong value after "let". Arrange that set A default {expr}; ... let A := A; will keep A at its current value when the value of expr changes. Before A is assigned a value by let (or in a data section), the value of A will change when expr changes value. Fix some memory faults that were possible after "fix" and "objective" commands in problems involving defined variables. Fix a bug that sometimes kept presolve from running (again in examples involving defined variables). 19940605: Fix storage allocation bug with "reset data". The bug sometimes caused subsequent unpredictable results. Fix bug with "restore" and defined variables: if "restore" caused variables declared after variables referenced by the defined variables to be instantiated, incorrect linear terms were generated, possibly causing a "solve_out bug" message. Example: var x{1..2} >= 0; var w; var y{1..1} = sum{i in 1..2} i*x[i]; minimize zot: 3*y[1]; foo: 4*x[1] + 5*x[2] = w; goo: 6*w = 1; drop foo; drop goo; solve; restore foo; restore goo; solve; # solve_out bug! Fix bug in handling recursive params and sets that have a := clause. The := values were not recomputed when values upon which they depend changed. Example: param p default 3; param q{i in 1..2} := if i == 1 then p else q[1]*p; display q; let p := 4; display q; # Did not change. Complain about "reset data foo;" and "let foo[...] := ...;" and data- mode assignments to foo if foo has a recursive := clause. 19940608: Fix some obscure glitches in error messages. 19940610: Fix bug in handling {if lexpr} notation (in an objective or a constraint declaration, when the lexpr is true). Example: var x{1..2} >= 0; minimize zot: x[1] + 2*x[2]; s.t. C{if 0 == 0}: x[1] + x[2] == 1; solve; # memory fault 19940613: Fix possible botch when there is an indexed collection of objectives and an "objective" command appears (e.g.) before all variables have had to be instantiated. 19940616: Fix glitch in flushing buffers after control-C (i.e., SIGINT) or after a nonzero return code from a solver. The glitch was only apparent with invocations like "ampl filename). The bug could cause a memory fault or other untoward behavior in shell and solve commands. Fix failure to correctly instantiate variables under complicated conditions involving "fix". Example: var x{1..2} := 1; var y{1..2} >= 0 := sqrt(.5); circle: sum{i in 1 .. 2} y[i]^2 = 1; var z = x[1]*x[2]^2 + 1; minimize zot: (z - 5)^2 + (x[1]-2)^2; drop circle; # now y is not needed solve; display x, z; fix y[1]; # Forces generation of y, which, in this # example, caused confusion with z. A simple # "fix y;" did not cause any problem. solve; display x, z; # Results should be the same as before # but were not. (In a more elaborate example, # the second solve gave a memory fault.) 19940829: Supply missing newline in "difference = ..." lines from presolve. The newline disappeared during the changes for version 19940808. 19940901: Fix bug in writing .nl files for problems in which one defined variable appears exclusively in a single network constraint (node declaration), and another appears exclusively in a general constraint. The bug caused solvers to suffer a memory fault. Example: var w; var x = 2 - w; node foo: net_in + x <= 1; var y; var z = y + 1; goo: z >= 2; 19940908: Fix a bug that could bite when linearization of piecewise-linear terms increases the number of variables beyond the next power of 2. Adjust option command and $ notation to treat options (environment variables) for which AMPL does not provide a default as described in the AMPL book. For example, option PATH; now will show the current PATH (used in Unix and MS-DOS systems for finding the programs invoked by the solve and shell commands), rather than saying "option PATH ''; #not defined", and if you change option PATH, then you can now restore it to its original value by saying option PATH $$PATH; Fix (rarely seen) bug related to common expressions that appear in entities, only some of which must be recomputed after a change. The fix could cause some changes in -T (option gentimes 1) output. Example: param p default 1; param n integer >= 0 default 1; var x{1..n} >= p-1; var x1{i in 1..n} = x[i]*p; var y >= p-1 <= 2; var y1 = y*p; zot: sum{i in 1..n}(x[i] + y1) = 1; solve; let n := n + 1; solve; # fault 19940914: Fix memory overwrite possible with the sequence 1. Do something that generates a variable. 2. Change something that causes the variable to be regenerated, but with an error (e.g., failed check or bad subscript) that interrupts the regeneration. 3. Correct the mistake and regenerate the variable. Fix bug in simplifying nonlinear sums of two terms, the second of which can have a unary minus operation "moved up". (The bug was only possible if presolve fixed at least one variable, so turning presolve off avoids the problem.) Example: param p default 0; var x := -1; var y = if x < 0 then x + 1 / (p - x) else x; var z; fix_z: z = 0; minimize zot: y; 19941003: Instantiate the entire index set of a for loop before starting execution of the loop, so the set of dummy indices for which the loop body is executed will be unaffected by assignments in the loop body. Example: set S default {1,2,3,4}; for {i in S} let S := S union {i + 4}; display S; # used to give 1..5; now gives 1..8 ## The loop could be stated more efficiently: ## let S := S union setof{i in S} i+4; Fix memory fault in "reset data p" when p has no data but has been displayed. Example: param p{i in 1..3} default i; display p; reset data p; # memory fault Fix glitch in handling missing ; after final end-of-file: in data mode, no error message appeared. Fix bug in discarding invalid subscripts: if a data section supplied values for a param or indexed collection of sets, and the associated set of subscripts duplicated a valid set, the latter was corrupted. This bug could only bite when the -o command-line option caused an instantiated problem to be written out, as in ampl -ogfoo foo Example: set J; set K; param a{J,K}; param b{K}; param c{J}; var x{J}; minimize zot: sum{k in K} b[k]*(sum{j in J} a[j,k]*x[j])^2 + sum{j in J} c[j]*x[j]; data; set J := 1 2 3; set K := 1 2 3 4; param a : 1 2 3 4 := 1 1.1 1.2 1.3 1.4 2 2.1 2.2 2.3 2.4 3 3.1 3.2 3.3 3.4 ; param b := 1 10.1 2 10.2 3 10.3 4 10.4 ; param c := 1 11.1 2 11.2 3 11.3 4 11.4 # Discarding this bad subscript made all ; # values of subscript [4] disappear. Fix an unlikely trap due to an uninitialized floating-point value being read as a signaling NaN (or denormalized number on systems with faulty hardware or systems software) in the course of a "let" command that assigns a value for the first time to a subscripted param. Allow write and solve commands to proceed if only error messages about discarded subscripts appear. Disallow write and solve commands when presolve complains about inconsistent bounds; at the second attempt, show the least value of $presolve_eps that would allow the command to proceed. Apply $presolve_fixeps test to the declared bounds on each variable; although the description of $presolve_fixeps suggested that it should apply to the declared bounds, it did not. Change to behavior of "model filename" and "data filename", which are now commands: AMPL returns to model mode at the end of the file unless the file ends in the middle of data. Change to behavior of "data" and (hitherto undocumented) "commands" commands: when they appear within a compound command (i.e., the body of a loop or the then or else part of an if command, or simply in a sequence of commands enclosed in braces), they are now executed when the flow of control reaches them, instead of when the compound command is being read. In this case, if they do not specify a file, AMPL reads commands or data from the current input file until it encounters either an "end" command or the end of the current file. New option insertprompt (default '<%d>'), if nonnull, specifies an insert-prompt (in which %d is the current insert level, i.e., nesting of "data" and "commands" commands specifying files and appearing within a compound command) that immediately precedes the usual prompt for input from stdin. New single-step mode, details of which may change: option single_step n; where n is a positive integer, specifies that if the insert level is at most n, AMPL should behave as though "commands -;" were inserted before each command: it should read commands from stdin until "end" or eof (control-D on Unix systems). Some special commands may appear in this mode: command meaning step execute the next command skip skip the next command next if the next command is an if-then-else or looping command, execute the entire compound command before stopping again (unless the compound command itself specifies "commands -;") cont execute until the end of the end of all currently nested compound commands at the current insert level Allow "reset data" and "update data" commands to appear in compound commands. New option solver_msg (default 0; called omit_solmsg until 19941007): if set to 0, the solution message normally printed by the solve and solution commands is suppressed. 19941005: Fix bug in handling repeat commands with omitted opt_until. Example: set A default {}; param p{A}; repeat {break;} display p; # memory fault Fix bug in { comandlist } within compound commands. Adjust some details of single-step mode. 19941007: Fix bug that caused default set expressions to be recomputed. Example: param p default 3; set S default p..p+2; display S; let p := 7; display S; # S was recomputed Change option omit_solmsg to solver_msg (default 1). 19941009: Fix glitch in "commands foo;" that kept commands following a compound command from executing. Diagnose premature end of file in "commands foo;". 19941028: Fix bugs in let commands assigning to a set or param that has a default expression when the set or param has not yet been assigned a value and appears on the right-hand side. Examples: param p{i in 1..3} default i^2; let{i in 3..3} p[3] := p[i] + 20; display p; # p[3] was 20 rather than 29 # did work correctly with p[i] changed to p[3] set T {1..4} default {}; let T[1] := T[1] union {"a"}; display T; # wrongly complained invalid subscript T[1] Omit (for now) recognition of @ (which once upon a time was a synonym for "cross" but now is reserved for future use). Adjust some insert-mode details; commands read by "include" or "model" are now at insert-level 0. Fix glitch in handling "option relax_integrality 1" when there are general (nonconvex resp. nonconcave) piecewise-linear terms. Fix bug leading to "tva top error" in the following example: param S; set A; param pmin{A} >= 0; param pmax{A} >= 0; param a{A}; param b{A}; param c{A}; # The consistency checks (>= ... and >...) on r # and L are necessary to elicit the bug. param r{i in A, s in 1..S} >= if s = 1 then 0 else r[i,s-1]; param L{i in A, s in 1..S-1} > if s = 1 then 0 else L[i,s-1]; data; param :A: pmin pmax a b c := a 2 16 12 2.4 3 b 5 25 103 3.4 2 c 1 5 24 1.5 5 d 10 54 105 1 5 e 1 5 1 5 0 f 0 13 16 2.4 3 g 0 22 107 3.4 2 h 0 5 28 1.5 5 i 0 51 109 1 5 j 0 5 1 5 0 ; param S := 6; for {i in A} # back to model/commands mode let r[i,1] := b[i]; ##The previous 2 lines would better be written ## let{i in A} r[i,1] := b[i]; ##but this did not reveal the bug. for {i in A, s in 1..S-1} { let L[i,s] := if pmin[i]=pmax[i] then pmin[i]+ .1*s else pmin[i] + (pmax[i]-pmin[i])/S * s; let r[i,s+1] := b[i] + 2*c[i]*L[i,s]; }; ##The above for loop would execute faster if written ## let{i in A, s in 1..S-1} L[i,s] := ... ## let{i in A, s in 1..S-1} r[i,s+1] := ... display r, L; New "read" command with syntax similar to the print command, except that the only simple or subscripted params, variables, and constraints (dual values) can be read. Optional input redirections are specified by < filename or < 'quoted_file_name' (before the read command's terminating semicolon). If no redirection is specified, values are read from the current input stream. To read from stdin, specify <- . Examples (reading from the current input steam): param p; read p; 4.2 display p; param ps symbolic; read ps; 'some funny text\ with a newline' display ps; param q{1..3}; read{i in 1..3} q[i]; 1.1 2.2 3.3 display q; param i; param j; param A{1..10,1..10}; param n; read n,{1..n}(i,j,A[i,j]); 3 2 4 2.4 3 5 3.5 4 7 4.7 display A; 19941030: Fix bug in reading sets in data mode after "update data": if two sets previously had the same value (and some command, such as "solve" or "display" gave occasion to check for sets having the same value), a data update to one changed the value of both, or (when both had default values) gave an erroneous error message about no data for the updated set. Examples: set A; set B; data;set A := ; set B := ; display A,B; update data A; data; set A := a b c; display A,B; # A = B (erroneously) reset; set A default {}; set B default {}; display A,B; updata data A; data; set A := a b c; display A,B; # erroneous message about no data for A 19941110: Arrange for $prompt2 to be shown for read commands inside general loop bodies. It was shown for for{i in 1..3} read p[i]; and the preferred read{i in 1..3} p[i]; but not for for{i in 1..3}{read p[i];} Fix glitch that sometimes caused "show" to mention set $display. Example: set A dimen 2; param p{A}; data; param :A: p := a b 1 b c 2 a c 3 ; option display_1col 1; display p; show; # sets: $display A 19941127: Omit check for violation of declared bound constraints by variables fixed with the "fix" command. Abort compound commands (and the reading of files other than stdin) after "Ignoring solve command..." or "Ignoring write command...". Fix memory fault in display of items with a declared index that is an ordered or circular subscripted set with no "by set" specification. Example: set A; set S{A} circular; param p{i in A, S[i]}; data; set A := a b; set S[a] := c d; set S[b] := d e; param p := a c 1 a d 2 b d 4 b e 5; display p; Fix bug with recursive symbolic parameters: if the value of such a param was another component of the same param, unpredictable behavior was sometimes possible. Example: set A circular; set B circular by A; param p{i in A} symbolic := if i in B then i else p[next(i)]; param q{B}; param r{i in A} := q[p[i]]; data; set A := A B C D E F G H I J K L M N O P Q R S T U V W X Y Z; set B := B F G J P Q V; param q := B 1 F 2 G 3 J 4 P 5 Q 6 V 7; display r; 19941205: Fix error handling common expressions in some sequences of commands, such as: param p in [0,10] default 7; param q in [0,p] default 4; set Z default 1..p; set W default 1..q; # p is common to q, W, and Z. let Z := 1..3; display Z,W; # faulted evaluating [0,p] for q. Fix bug in sets involving random functions used in contexts that only involve membership tests. Example: set ZL := {1..5: Uniform01() < .3}; param p{i in 1..5} := if i in ZL then 0 else Uniform01() + 1; display p; # only requires testing membership in ZL display ZL; # forces ZL to be fully generated Fix bug in handling file redirections introduced with the read command. Unpredictable results could result from a memory-overwrite bug associated with redirections. Fix bug (introduced 28 Aug 1994?) in generating variables (without suitable data changes) that appear syntactically before previously generated variables. The bug could cause objectives or constraint bodies to be miscomputed. Example: var unused >= 0; var x{1..2} >= 0; convex: x[1] + x[2] = 1; var v; vdef: v = x[1] + 2*x[2]; minimize zot: v; solve; print unused; # appears syntatically before the other vars print zot; # wrong value 19941209: Fix bug in handling variables used as params (not used in any constraints or objectives). Instantiating them after a "solve" could cause a fault in a second "solve" if nothing has intervened to suitably change the problem. Example: var x{1..2} >= 0; var y; minimize zot: sum{i in 1..2} i*x[i]; s.t. convex: sum{i in 1..2} x[i] = 1; solve; let y := sum{i in 1..2} i*x[i]; print y, zot; solve; # fault Fix bug in handling repeat loops that appear at the end of the last input file and do not end with a semicolon. In one lengthy example, an infinite loop resulted. 19941212: Fix infinite loop in linearizing piecewise-linear terms for problems with both "easy" terms (not requiring integer programming) and at least one "hard" term in a constraint eliminated by presolve. Example: var x{1..2} >= 0; var y >= 0; minimize zot: sum{i in 1..2} i*x[i] + <<.7;3,4>>x[1]; s.t. convex: sum{i in 1..2} x[i] = 1; s.t. bletch: <<.5;1,2>>x[1] >= y; s.t. fix_y: y = 0; # presolve deduces that bletch always holds 19941220: Fix bug (e.g., memory fault) in regenerating variables after entities in their declarations, such as their index sets, have changed and the variables have participated in commands that do not require instantiating the entire current problem (i.e., most commands other than solution, solve, or write, unless dot notation forces the problem to be instantiated). Example: set A ordered; var x{A} >= 0 <= 10; minimize zip: sum{i in A} ord(i)*x[i]; var y{i in A: i != last(A)} = x[i] + x[next(i)]; s.t. zot{i in A: i != last(A)}: y[i] = 1; data; set A := a b c d; solve; reset data A; data; set A := g h i j; drop{i in {'g','h'}} zot[i]; solve; # fault Fix memory-overwrite bug with "update data": the sequence data; /* read one or more indexed params */ update data; # no use made yet of the params just read data; /* read more of the same params */ caused unpredictable behavior. 19950101: Fix bug that crept in sometime in or about fall, 1994: after data changes (e.g., "reset data" or "let"), objectives or constraints used with column-generation (including arcs) were sometimes miscomputed. Example: minimize cost; param p{i in 1..2} default i; var x{i in 1..2} >= 0 obj cost p[i]; convex: sum{i in 1..2} x[i] = 1; solve; # objective 1 (correctly) reset data; solve; # objective 2 (wrong! should be 1) 19950204: Fix bug in handling changes to option substout. Allow UTF-encoded Unicode in option names. Fix glitch in reading UTF in models on machines where chars are signed by default. Fix memory allocation bug that could appear after a message about discarding bad subscripts for a subscripted param (or set) that was given a single value in a data section, as in set A; param p{A}; data; set A := abc; param p := xyz 3; display p; Fix two bugs with option linelim 1: 1. Nonlinear defined variables involving an additive constant (nonlinear expression + constant) had the constant added twice to constraints, objectives, and other defined-variable values linearly involving the defined variable. Example (before bug fix): ampl: var x; ampl: var y = x^2 + 1; ampl: var z = 3*y + 4; ampl: minimize zot: y + z; ampl: option linelim 1; ampl: solve; MINOS 5.4: optimal solution found. 2 iterations, objective 19 # should be 8 ampl: display x,y,z,zot; x = -5.42101e-20 y = 1 z = 10 # should be 7 zot = 11 # should be 8 2. In a problem with nonlinear defined variables involving some variables linearly and only used linearly, the linear contributions to the nonlinear defined variable were added in twice. Example (before bug fix): ampl: var x; ampl: var y = x^2 - 2*x; ampl: var z = 3*y; ampl: minimize foo: y + z; ampl: option linelim 1; ampl: solve; MINOS 5.4: optimal solution found. 2 iterations, objective -30.25 # should be -4 at x = 1 ampl: display x,y,z,foo; x = 2.75 y = 2.0625 z = -10.3125 # should be 6.1875 (for x = 2.75): low by 6*2.75 foo = -8.25 # should be +8.25 (for x = 2.75) (Without a previous "solve", "solution", or "write", the display commands in the last two examples computed correct values from the original expressions.) 19950209: Fix a bug (introduced in 19941028) that could cause a memory fault in or after a "close" command. (With the bug, "close" commands are often harmless, which is why we didn't see the bug until now...) 19950211: Fix some memory leaks (visible with repeated "reset data" commands). Fix glitch that caused for {k in 1..10} commands foo; to read from stdin after reading foo (on every iteration). 19950307: Fix bug in handling partially dropped defining constraints. 19950315: Provisional versions (subject to change as we gain experience and get feedback) of several extensions are now available. New reserved word "all", used in "drop all;", "fix all;", "restore all;", "unfix all;". Extension to "objective" command: "objective;" or "objective >filename;" prints commands establishing the current drop status of objectives. In particular, if one had previously said "objective foo[3];", "objective;" would print "objective foo[3];". Similarly, "drop;" or "drop >filename;" or "restore;" or "restore > filename;" prints commands establishing the current drop state of the constraints and objectives, and "fix;" or "fix >filename;" or "unfix;" or "unfix >filename;" prints commands establishing the current "fix" state of the variables. In these contexts, "drop" and "restore" are interchangeable, as are "fix" and "unfix". New "problem" declaration/command has three functions: declaring a new problem, making a previously declared problem current, and printing the name of the current problem (in the form of a problem command establishing the current problem). problem name optional_indexing optional_environ : itemlist ; declares a new problem and specifies the variables, constraints, and objectives that are in it. Other variables appearing in the specified constraints and objectives are fixed (but can be unfixed by the "unfix" command). The new problem becomes the current problem. Initially the current problem is "Initial". The "itemlist" in a problem declaration is a comma-separated list of possibly subscripted names of variables, constraints, and objectives, each optionally preceded by an indexing, as in {i in A} foo[i]. More generally, nested indexings similar to those allowed in function calls may be specified, as in {i in A} (foo[i], goo[i], {(i,j) in B} zoo[i,j]) The command problem name; makes name (a previously declared problem) current. And problem; or problem >filename; prints the current problem name (as "problem name;"). Drop/restore and fix/unfix commands apply only to the current problem. Variable values, like params, are global; just the fixed/unfixed status of a variable depends on the problem. Similarly, the drop/restore status of a constraint depends on the problem (as do reduced costs). The current problem does not restrict the "let" command. When a problem is declared, it can optionally specify an environment associated with the problem: the optional_environ phrase has the form environ envname to specify that the problem's initial environment is envname. Otherwise a new environment with the same name as the problem is created, and it inherits the then current environment (set of option values). In option commands, unadorned (conventional) option names refer to options in the current environment, and the notation envname.opname refers to $opname in environment envname. The new declaration environ envname optional_indexing; declares a environment envname (or a set of environments, subscripted by the indexing if specified). If there is no indexing, envname becomes the current environment for the current problem. New command environ optional_indexing envname := envname1; where envname and envname1 are optionally subscripted environment names, copies environment envname1 to envname. New "expand" command prints generated constraints and objectives (much as the linrc program does): expand [indexing] itemlist [>filename]; The itemlist can assume the same forms allowed in problem declarations. If it is empty, all non-dropped constraints and objectives are expanded. The variant solexpand [indexing] itemlist [>filename]; shows how constraints and objectives appear to the solver. It omits constraints and variables eliminated by presolve unless they are explicitly specified in the itemlist. Both the "expand" and "solexpand" commands permit variables to appear in the itemlist; for each, the commands show the linear coefficients of the variable in the relevant (non-dropped and, for solexpand, not eliminated by presolve) constraints and objectives, and indicates " + nonlinear" when the variable also participates nonlinearly in a constraint or objective. New options expand_precision and expand_round control printing of numbers by expand. By default they are currently printed to 6 significant figures (option expand_precision 6). Change to function calling conventions: symbolic arguments were formerly quoted (as though they were symbols in a data section); now they are stripped of quotes and the \ before a newline. Examples: print 'a b'; now prints a b rather than 'a b' New printf format %q prints with data-section quoting rules (omit quotes if omitting them is allowed in a data section); new printf format %Q always quotes strings. New concatenation operator & has precedence below all arithmetic operators and performs string concatenation. It accepts numeric operands and converts them to full-precision decimal strings (as though by printf format "%.g": recall that AMPL's "%.0g" gives full precision, rather than behaving like "%.1g"). Contexts (other than alias strings in declarations) that previously required literal strings now also accept an expression in parentheses. Expressions in commands may involve operands of the form $value (a $ followed by an environment name) and $environ.value (where environ is the possibly subscripted and previously declared name of an environment). $values may not be used in declarations. New builtin functions involved with strings: num('12.34') = 12.34 # convert string to number num('12.34x') = error # complain if stripping leading and # trailing white space doesn't yield # a valid decimal number num0('12.34x') = 12.34 # strip leading white space, and # interpret as much as possible as # a number, but never complain ichar('A') = 65 # Unicode value of the first character # in its argument string char(65) = 'A' # inverse of ichar length('abcd') = 4 # length of string substr('abcdef',3,2) = 'cd' # substring substr('abcdef',3) = 'cdef' # substring sprintf("stuff %.2f blah %g Blah %.g", 13/3, 2/7, 3/11) = 'stuff 4.33 blah 0.285714 Blah 0.2727272727272727' # general formatted conversion to string match('abcde','cd') = 3 # starting position of arg2 in arg1 match('abcde','xy') = 0 # or 0 if not found; arg2 can be a general # regular expression sub('abcdecd','cd','XYZ') = 'abXYZecd' # substitute arg3 for the first occurrence # of arg2 in arg1 gsub('abcdecd','cd','XYZ') = 'abXYZeXYZ' # substitute arg3 for all occurrences of # of arg2 in arg1 arity('S') = arity of S if S is a set; else 0 # for use with _SETS, as in # display{s in _SETS} arith(s); There is no implicit conversion of strings to numbers, but new builtin functions num(string) and num0(string) perform explicit conversions. Both ignore leading and trailing white space; num complains if what remains is not a valid number, whereas num0 just converts as much as it can (returning 0 if it sees no digits). The expressions 'abc' & x+3 and 'abc' & sprintf("%.g",x+3) yield the same strings. Now, e.g., setof{i in 1..3} 'ABC' & i = {'ABC1', 'ABC2', 'ABC3'}. The match, sub, and gsub functions accept strings representing regular expressions as their second arguments. Such expressions are as in plan 9. They are similar to the regular expressions recognized by the Unix editors ed and sed, except that parentheses as operators must not be escaped, and, in addition to * for 0 or more occurrences of the preceding item, + means 1 or more occurrences, and ? means 0 or 1 occurrence. The replacement patterns (third arguments) for sub and gsub are like those for ed and sed: & stands for the whole matched string, as does \0, and \1, \2, ... \9 stand for the string matched by the first, second, ..., ninth parenthesized expression in the pattern. New "cd" command reports or changes the current working directory. New automatically updated params _nvars = number of variables in current model _ncons = number of constraints in " " _nobjs = number of objectives in " " _varname{1.._nvars} = names of variables in current model _conname{1.._ncons} = " " " " " " _objname{1.._nobjs} = " " " " " " and synonyms for current model entities: _var{1.._nvars} = synonyms for variables in current model _con{1.._ncons} = " " constraints " " " _obj{1.._nobjs} = " " objectives " " " These synonyms can be used in display and other commands. They present the modeler's view (before presolve). Similarly automatically updated entities with _ changed to _s (i.e., _snvars, _svarnames, _svar, etc.) give the solver's view, i.e., the view after presolve. New automatically updated sets: _PARS = set of all declared param names _SETS = " " " " set " _VARS = " " " " variable " _CONS = " " " " constraint names _OBJS = " " " " objective " _PROBS = " " " " problem " _ENVS = " " " " environment " _FUNCS = " " " " (user-defined) functions These sets may appear in commands, such as display and print. New additions to the reserved-word list: Current, Initial, all, environ, option. The other new names appearing above may be declared to be something else. 19950317: Fix memory fault in single-stepping when data turn up missing after an immediate command. Fix bugs in subscripted references to _varname, _conname, _objname, and to dot notation with subscripted _var, _con, _obj. 19950412: Recover somewhat better from floating-point exceptions. Fix glitch that prevented overriding builtin function names with problem names. Fix bug in an unadvertised check statement syntax: check {if lexpr}: lexpr1; Such checks appeared never to be tested. Now they are enforced. Fix bug that caused var x := 2; print x; to say "ungenerated variable x". Enforce the restriction that $values and synonyms (such as _VARS and _nvars) may not appear in declarations. Fix glitch that caused the script param p; var x; minimize zot: (x-p)^2; for{i in 1..2}{ let p := i; solve; display x, x.rc; } to give Error executing "let" command error processing objective zot: no value for p The problem was in handling dot notations (like x.rc) that require instantiating a model; the script ran correctly with the param p declaration changed to "param p default 0;". Abort reading commands from files other than stdin after detecting an error. Fix glitch that prevented the command following a command referencing an undeclared entity from being executed (when reading commands from stdin). For example, print c; # with c not declared print 3; # didn't execute Fix bugs in the Gamma and Poisson random function (a failure to retain values across calls that could cause them to compute wrong values or memory fault). Fix bugs in reading dual variable values for partially dropped constraints from .sol files (in the solve or solution commands), and in evaluating expressions involving dual values of dropped constraints. Fix glitch that (on some systems) could cause the error message "adjoin: SBUFLEN (bug!) is too small". Fix possible memory fault with single-stepping of break and continue commands. WATCOM binaries are now compiled to circumvent the Pentium divide bug. New xref command tells which entities depend directly or indirectly on specified item. Syntax: xref item_list opt_redirect; where item_list is a (white-space or comma separated) list of model entities and opt_redirect is an optional file redirection. Preliminary (and possibly still buggy) implementations of xref, delete, purge, and redeclare: New command xref shows entities that depend directly or indirectly on specified entities. Syntax: xref itemlist optional_redirection; New command delete foo; deletes foo, restoring any previous meaning foo had, provided no other entities depend on foo, i.e., if "xref foo;" reports no dependents. New command purge foo; deletes foo and all its (direct or indirect) dependents. New form of declaration: "redeclare" followed by an ordinary redeclaration replaces any existing declaration of the specified entity with the given one, provided either that the entity has no dependents, or that the new declaration does not change the character of the entity (its kind, such as set or param, and its number of subscripts). Redeclarations that would cause circular dependencies are rejected. New symbolic system parameter solve_message is assigned the message shown (when not suppressed by "option solver_msg 0;") by the solve and solution commands. One can assign solve_message with "let", but may not delete or redeclare it. New variants of break: break commands; break all; terminate, respectively, the current commands command or all commands commands, if any, and otherwise act as a "quit" command. 19950416: Fix memory fault caused by evaluating variable.val before some command such as "solve" forces presolve to run. Fix bug in switching to a new environment with values that are set in the current environment but unset in the new one. Fix a bug (memory overwrite) in handling subscripted problems. Fix bug introduced in version 19950314 that made $solver_oopt (specification of output format for $solver) and $solver_auxfiles misbehave. Fix bug in handling subscripted objectives with differing constant terms: if some elements were dropped, the remaining (nondropped) ones got the wrong constant term (in the .nl file). This only affected objective values reported by solvers. Fix glitches in expand command that suppressed variables fixed by presolve and constraints removed by presolve; that was only supposed to happen for solexpand. Adjust the delete and purge commands to act on more than the first element in their itemlists. Allow items of the form "option opname" in the itemlist for delete. This removes opname from the current environment. Adjust "reset;" command to terminate all "commands" and "data" commands: "reset;" is an immediate command, executed when seen. Previously, it appeared to work when executed in single-step mode, but a commands-reading procedure was left running, and chaos could ensue when it terminated. Extend $solver_auxfiles interpretation: capital letters have no affect on linear problems, but on nonlinear problems (including problems with binary or integer variables) are treated the same as their lower-case equivalents. The default $minos_auxfiles is now RF rather than rf, causing .row and .fix files to be written only for nonlinear problems. 19950501: Fix memory fault in redeclaring objectives. New function indexarity('foo') gives the arity of foo's index set if foo has been declared to be something indexed, or 0 if foo has been declared as something that is not indexed, or -1 if foo has not been declared. Example: display{i in _PARS} indexarity(i); 19950517: Quote symbolic subscripts in more error messages. Expand command: try to keep number*variable on the same line (so far just for linear terms). Fix bug in handling nonlinear defined variables under -L (option linelim 1) when they involve linear terms and appear linearly in constraints, objectives, or other defined variables. Example: var x; var y = x^2 - 2*x + 1; minimize zot: 2*x + 3*y; option linelim 1, nl_comments 1; write gf; # bogus v number for 3*y (in f.nl) Fixing this bug led to an improvement in the classification of nonlinear variables sometimes changes the resulting .nl file (and could slightly reduce the work some solvers do). It also revealed a long-standing bug in the AMPL/solver interface that caused wrong derivatives to be computed in problems that involve defined variables with right-hand sides of the form another_defined_variable + linear_expression Fix bug in printing a hitherto unused variable after a solve. Example: var x{1..2} >= 0; var y := 42; convex: x[1] + x[2] = 1; minimize zot: x[1] + 2*x[2]; solve; display y; #bogus value Check statements are now only enforced during solve, write, and solution commands, or when the new command check; is executed. Dot notations that require presolve are now excluded from declarations. 19950519: Fix horrid, long-standing bug with subscripted sets of arity > 1 (apparently introduced in version 19940128): when such sets had more than 32 members, tests of membership in them went awry. Example: set A; set B{A} dimen 2; set C dimen 2; set D := {i in A, (j,k) in C diff B[i]}; data; set A := a b; set B[a] := (x,x) (x,y); set B[b] := (y,x) (y,y); set C := (x,x) (w,x) (y,y) (y,z); display D; # correct let B['a'] := B['a'] union {1..4,1..8}; # now B['a'] has more than 32 members display D; # botched: (a,w,x) was erroneously in D 19950521: New variant of close command: close; closes all files opened by redirections. 19950524: Fix bug in presolve's handling of piecewise-linear terms that can be turned into linear terms because of deduced variable bounds: constraints (and defined variables) specifying a variable to be the sum of such terms where mishandled. Break out of compound commands when the solution, write, solve commands fail. 19950530: Fix bug (possible memory fault or at least botched error message) in the error message xxx does not work with variables for xxx = "match", "OPNUM2STR". Fix possible glitch in the expand commands printing of nonlinear expressions (a length value wasn't returned, which could cause some lines to be too long or too short). Fix memory fault with multiple occurrences of the same $env[something].val expression (under the right circumstances). Example: set A default {'a'}; environ E{A}; option E['a'].solver cplex; print {a in A}: $E[a].solver, $E[a].solver; # fault Fix bug in testing membership in Cartesian products (under the right conditions). Depending on the compiler, the test may have come out wrong. 19950608 Fix bug in "reset data;": storage for subscripted items whose actual index set has only a single member (e.g., which are given a single value in a data section), and for subscripted sets in general, went onto the wrong free-list, possibly causing unpredictable subsequent behavior. 19950614 Fix bugs in handling strings in the ord, next, and prev functions. Example: set A ordered := {'a', 'b', 'c'}; print ord('b',A); # fault print next('b',A); # fault Arrange for option funcwarn only to affect constraint and objective declarations. Thus it no longer affects function zot; param p := zot(3); # used to get msg about variable in := display p; # now elicits msg about zot missing Change in writing of auxiliary .fix file: values of variables fixed by presolve are suppressed unless $auxfiles or $($solver & '_auxfiles') contains "v". (The primary reason use of the .fix file is now to convey names of defined variables for use in printing error messages. This change omits sometimes lengthy work that is hardly ever useful.) Fix glitch in dropping nodes: if something forced a dropped node to be regenerated (because of its appearance in arc declarations), incorrect constraint names might sometimes appear, unpredictable behavior might occur because of a memory-overwrite error, and if the problem had defined variables, an infinite loop resulted (in, e.g., solve and write commands). Fix bug in handling set expressions in read commands: expressions within the set expressions were sometimes not computed, which could cause a memory fault or other untoward behavior. Example: param p default 2; param q{1..p}; read{i in 1..p} q[i]; # memory fault 1.1 2.2 display q; Warn of missing subscripts on sets. (Later we hope to warn about other missing subscripts, but that's harder.) 19950618 Fix glitch with "system sets" _PARS, _SETS, etc.: displaying two at once led to "unexpected subtype 16 in same_set". 19950620 Fix long-standing bug in handling if expressions in a nonlinear context when the test can be moved outside of a loop. Example: ampl: var x{1..2,1..2} >= 0;; ampl: minimize zot: sum{i in 1..2, j in 1..2} ampl? sin((if i == 2 then 10+j else j)*x[i,j]); ampl: solve; error processing objective zot: unexpected rewclass 4 for op 0x14e2 in ecopy 19950716 Fix bug in redeclarations of, e.g., indexed params with default expressions. Something got freed twice, causing unpredictable subsequent behavior. Increase by a factor of 256 the largest allowed single memory allocation. This increases the number of subscripts (members in indexing set) for an indexed param from 4194304 to 1073741824. Banish the message "More than 2000 columns; must increase CNMAX." In redeclarations that do not specify a := value, retain data where possible. Fix bug in deleting for and repeat loops: if the loops' controlling expressions involved constants (such as 0) or expressions involving only constants (such as atan(1)) that appeared in the model, subsequent attempts to solve the model could lead to unpredictable behavior. Fix bug (e.g., memory fault) in evaluating objective values and constraint bodies after an error interrupts a solve or write command. Fix memory-overwrite bug in initially dropping a constraint, changing something (e.g., a set or param) on which the constraint depends, then restoring and trying to use the constraint. (The change to the ungenerated constraint's prerequisites caused an erroneous attempt to save the constraint's dual variables before generating the constraint.) 19950619 Introduce variant _display of the display command; _display is meant for possible use by front ends. Like the display command, _display emits one or more tables, but _display replaces display's table header with a line consisting of "_display" and three integers: s = number of subscripts of each item in each table line k = number of items on each table line n = number of table lines Each table line consists of s subscripts followed by k items, all separated by commas. No semicolon is appended to the table. 19950720 Fix memory fault in "write;" (which is supposed to be the same as "write $outopt;"). Fix bug in simplifying constant expressions in a nonlinear context. Example: set S; set T; var x{T} >= 0; param p{S,T}; minimize zot: sum{i in S} exp(sum{j in T} p[i,j]*(p[i,j]+1)*x[j]); data; set S := a b; set T := d e; param p := [a,*] d 0 e -1 [b,*] d 3 e 4; # For i = 'a', exp(sum{j in T} p[i,j]*(p[i,j]+1)*x[j]) # vanishes; a bug caused a memory fault during simplification. Add another variant of the display command: csvdisplay is similar to _display, except that it only writes one table (complaining if asked to write more than one), and it replaces the initial _display output that describes the following table with $1,$2,...,$n (where n is the number of items in one line of the table that follows). New options csvdisplay_precision and csvdisplay_round govern the precision of printing for csvdisplay; the defaults (0 and '') give full precision. 19950726 Adjust _display's printing so numeric precision is governed by $csvdisplay_precision and $csvdisplay_round. In -M output and printing by the show command, insert "s.t. " before constraints. New option csvdisplay_header (default 1): if nonzero, cvsdisplay's first line now has the form Index_1,Index_2,...,Index_k,Expr_1,...,Expr_n (where k is the number of subscripts and n the number of expressions in one line of the table that follows); if the jth expression is a simple name (not of the form Index_i or Expr_i) that has not appeared previously in the current csvdisplay header line, then that name appears rather than "Expr_j". If $csvdisplay_header is 0, this header line is omitted. Adjust line breaking for "show" and omit some extraneous commas. Diagnose some more undefined names (e.g., in problem declarations). Recognize repetition counts in next nnn skip nnn step nnn Fix bug (possible memory corruption) in handling solver messages over 120 characters long. Fix failure to complain about invalid subscripts in set A; set B{A}; data; set A := a b c; set B[a] := ab bc; set B[b] := ba bb; display B; let A := {}; display B; # didn't complain about bad B subscripts Fix bugs introduced in 19950315 in handling option commands with old-style concatenations as values. Examples: option solver $solver b; # should append b to $solver, # but gave syntax error option solver $solver 'b'; # same: but it looped. Fix bug in handling "reset data; data diet.dat; solution diet.sol;" before reading any data: the solution command appeared to be ignored. 19950727 Adjust show command to omit $display from the list of declared sets. Example: display{i in 1..5,j in 5..10} 10*i+j; show s; # erroneously mentioned $display 19950801 Fix glitches in -MD and -O introduced in the last changes (19950727). Do not break compound commands or command-line file processing when presolve determines the solution (e.g., eliminates or fixes all variables). Fix glitch in parsing if $solver == $$solver then (Workaround: put parens around the logical expression.) Fix possible memory fault associated with the match and sub functions. 19950802 Fix glitch with csvdisplay's display of 2-D tables with option display_1col 0, omit_zero_rows 1 /* small value */; Adjust "show" to reflect current drop/restore status after a problem command. 19950804 Fix bug in saving problem drop/restore (and fix/unfix) status: changes away from the status declared for the problem made the first time the problem is current were lost. Example: # First two problem declarations problem Sub: Artif_Reduced_Cost, Trans, Supply, Demand; problem Master: Artificial, Weight, Excess, Multi, Convex; # Master is current; change away from its declaration: drop Artificial; restore Total_Cost; fix Excess; # Do something with another problem problem Sub; drop Artif_Reduced_Cost; restore Reduced_Cost; # Switch problems: changes away from problem declaration # were lost. problem Master; show obj; show var; Inhibit executing commands in files other than stdin after the message "Abandoning compound command to accept ...". 19950808 Fix bug with -P (option presolve 0): if declared bounds caused a piecewise-linear term in a constraint to turn into a linear term, memory could be corrupted in an unpredictable way (leading to subsequent chaos). Fix glitch in show command: in the following example, the show command gave "(o.0 := avail)" rather than "avail" as the right- hand side. model steel.mod; data steel.dat; print total_profit, avail; show Time; 19950810 Fix bug in "expand": piecewise-linear terms after the first one printed had the wrong slope and breakpoint lists: they duplicated those of the first (in each constraint or objective). Allow the expand and solexpand commands to show dropped families of constraints. Previously, the sequence set A; s.t. foo{A}: /* ... */; drop foo; expand foo; resulted in no output; now it shows the foo constraints, adorned with the comment "# Dropped". Fix bug in solexpand's test for constraints eliminated by presolve. Adjust solexpand to say "# Presolve completely eliminated foo." in response to "solexpand foo;" when appropriate. Previously the solexpand command gave no output in this case. Add the comment "# Eliminated by presolve." when appropriate to solexpand output for a subscripted constraint. Thus solexpand foo[3]; might elicit this comment, whereas solexpand foo; might elicit the comment "# Presolve completely eliminated foo." Correct a possible memory fault with the show command (in complicated situations). Allow the print command to have no arguments, which causes it to behave the same as printf "\n"; 19950811 Fix bug in reading dual values for problems with free rows when -P (option presolve 0) is in effect. Presolve eliminates free rows (i.e., constraints of the form -Infinity <= stuff <= Infinity), but -P allows them to be conveyed to the solver. The bug could result in incorrect dual values for some constraints, and could cause memory overwrites and subsequent unpredictable behavior. 19950815 Fix possible memory fault in the cd command with WATCOM-compiled binaries. Fix botch in deleting subscripted set references in command contexts other than expressions appearing in an indexing expression (memory for which is not yet recovered). Subsequent commands referencing the subscripted set could give unpredictable behavior. Example: set A; set B{A}; data; set A := a b c; set B[a] := x; set B[b] := ; set B[c] := x; for{i in A} printf "'x' is%s in B[%s]\n", if 'x' in B[i] then '' else ' not', i; # deletion of ('x' in B[i]) was botched, giving different # results to a repeat of the same command: for{i in A} printf "'x' is%s in B[%s]\n", if 'x' in B[i] then '' else ' not', i; 19950827 Fix bug in handling in handling string-valued loop invariants, as in set A; param p{A} symbolic; set B; param q{B} symbolic; for{i in A, j in B} { if p[i] == q[j] then print i,j; } (in which p[i] is invariant as j ranges over B). If it bit, the bug could cause unpredictable behavior. Fix a bug that could give the error message "presolve vT bug". The bug sometimes bit after presolve deduced sharper bounds that turned a piecewise-linear term into a linear term. Fix a bug in handling complicated column-generation structures after, e.g., a data update followed by a printing command that caused some (but not all) entities necessary for "solve" to be updated. An ensuing solve sent the wrong problem to the solver. (The bug fix entails a significant change to the logic for deciding what needs to be recomputed.) Fix a memory fault in executing "expand undefined_name;". Fix a bug with renumbering defined variables after some change forces "earlier" variables to be regenerated (or generated for the first time). The bug could cause the message "solve_out bug" (or worse, could silently cause the wrong problem to be generated). Fix an infinite loop in deleting a sum that expanded into the sum of two variables. Fix glitch in displaying a variable indexed over an empty set when the variable has no := or default clause and is not given an (empty) initial value in a data section. Put parens around names generated for variables added to linearize piecewise-linear terms. Fix glitch in recomputing some set expressions in for loops. New debugging option: $debug_stub, if not '' (its default value), is used to construct stubs of the form ($debug_stub & '_1'), ($debug_stub & '_2'), ... in solve commands. If a file named ($debug_stub & '_1.sol'), ($debug_stub & '_2.sol'), etc. exists, "solve" behaves as "solution" and reads the .sol file. Otherwise it leaves behind its .nl and .sol files, so a subsequent ampl run can read the .sol files (and the .nl files are available for debugging purposes). This permits recreating some situations without rerunning the solvers involved. The .nl files (and hence .sol) files are ASCII files to make it easy to create them on one kind of machine and use them on another. 19951009 Fix glitch that forced one to use .val notation when referring to the value of a variable in the set specification of a for loop. Fix bug with drop scalar_constraint; solve; restore scalar_constraint; expand; (where scalar_constraint is an unsubscripted constraint): the "expand;" did not show scalar_constraint. Fix bug that caused set A dimen 2 default {}; var x{1..3}; s.t. c{(i,j) in A}: sum{k in 1..3} i*x[k] >= j; display c; to give the error message "No values for c.". Now it says "c; #empty". Fix botch in handling slices of the same kind over different sets when the sets turn out to be the same (have the same members): because of erroneously freed memory, unpredictable behavior was possible. Fix redeclare bug revealed by: var x {0..10}; minimize obj: sum {j in 1..10} x[j]; subj to jump {j in 1..10}: x[j-1] = x[j] - 1; solve; redeclare subj to jump {j in 1..10}: x[j] = x[j-1] + 1; solve; #fault Fix a bug with arcs as defined variables: sometimes they gave rise to a constraint with an unutterable name, rather than a nonlinear defined variable. Now they act as other defined variables. Fix bug with "delete" and "purge" that could cause subsequent "show" commands to fault or print the wrong entity. New command: reset function; closes all pipe functions and unlinks all dynamically linked functions, causing them to be restarted or relinked if invoked again. For a specific function foo, reset data foo; acts just on function foo. New command: delete check n; deletes the n-th check. Extension of redeclare syntax: redeclare check n optional_indexing : ...; redeclares the n-th check. New addition to reserved-word list: dotname (for use in a forthcoming extension). 19951010 Fix botch with "let" set assignments, revealed by set S ordered; let S := {"A","B","C","D"}; display {i in S} ord(i); # A had ord 4 rather than 1 19951015 Fix bug in computing .dual, .lb1, .lb2, .lbs, .ub1, .ub2, .lbs, .ubs notations for partially dropped constraints. 19951019 Fix buglet with "fix;" that caused it to show defined variables (as unfixed). Fix bug in handling indexed params (etc.) that have a default expression. If "let" or "read" or "update data" caused the param to have data for invalid subscripts, causing an error message about the data being discarded, and some subsequent command caused the default expression to be evaluated, chaos could ensue. Fix bug in handling problem declarations with indexing expressions: the indexing expressions were not necessarily reevaluated after "let" or "read" or "update data" should have affected them. Fix bug with "commands foo.dat;" when foo.dat consists of or ends with a data section that starts with "data;". 19951020 Allow "reset data p;" even when p is declared with a := expression, to force recomputation of random functions in the := expression (and to force reevaluation of any user-defined functions in the expression). Extend "reset data;" to force recomputation of all := expressions. 19951027 Fix bug with "show" after "delete": param p; param q; show par; delete p; show par; said "parameter: p". Fix bug in deleting objectives (with the delete or purge command): an expression got deleted twice, which could cause subsequent chaos. Fix memory allocation bug in discarding subscripted problems when the problem's indexing set shrinks. Fix another bug in handling problem declarations with indexing -- another context where the indexing expressions were not necessarily reevaluated after "let" or "read" or "update data" should have affected them. Now problems (including the current one) should be adjusted when their indexing expressions change, except that previous explicit drop, restore, fix, and unfix commands remain in effect. The new "reset problem" command cancels this treatment of previous explicit drop, restore, fix, and unfix commands, and brings the problem to its declared drop/fix state. This command has the forms reset problem; # applies to the current problem reset problem foo; reset problem goo[subscripting]; If the latter forms mention the current problem, they have the same effect as the first form. For now, at least, "reset problem" does not affect the problem's environment. 19951130 Fix precedence bug in printing nonlinear expressions involving sums: parentheses were omitted incorrectly. Example: var x; minimize zot: (x + 2)^2 + x*(x^2 + 3*x); expand; gave minimize zot: 2 + x^2 + x*x^2 + 3*x; Adjust error reporting for discarded subscripts so this "error" won't cause compound commands to be aborted, or "solve" to think that presolve found the problem infeasible. Fix bug in handling slices: chaos could sometimes ensue if, e.g., one tried to compute with a set for which no data had been given. 19951202 Fix bug in evaluating a dropped objective before anything else requires the collect phase. Example: var x; minimize zip: (x-1)^2; minimize zap: (x-2)^2; objective zap; print zip; Fix bug introduced 19951027 in the above example with "objective zap" changed to "drop zip". Arrange for option relax_integrality 2; to assume integrality during presolve, but tell the solver that there are no integer variables. Force presolve to run again after $relax_integrality has changed. Increase default value of $pl_bigM from 1e4 to 1e6. Arrange for changes to $pl_bigM to cause presolve to run again when its results would be affected by the change to $pl_bigM. 19951204 Fix memory fault (or worse) with "delete" after "redeclare". Example: param p; redeclare set p; delete p; display p; # fault Arrange to run presolve (when nothing else has caused it to run) so defined variables involved in dropped objectives or constraints appear to have the "right" values. Example: var x := 2; var y = x^2 + 1; minimize zot: y + 10; drop zot; print zot; # used to print 11 rather than 15 Disallow declarations involving defined_variable.val. (Instead, declare a new param, use the param in the declaration, and use "let" to give it the right value. This helps clarify when things are to be evaluated.) 19951219 Fix bug with deleting or redeclaring a variable, constraint, or objective or an entity declared before a variable, constraint, or objective. Fix memory fault with "update data" after deleting something, but not everything declared after it. Fix bug with redeclaring a param with a recursive := clause. 19960101 Fix a glitch in processing var = declarations in which the same variable appears linearly two or more times in the right-hand side. The glitch only affects a forthcoming extension to the AMPL/solver interface for recognizing partially separable structure and computing Hessians. 19960104 Fix bug in handling slices of sets declared with := when the right-hand side changes. Example: set A; param p{A}; set B := setof{i in A} (i,p[i]); data; param :A: p := 1 11 2 12 3 13; display {(i,12) in B}; reset data; data; param :A: p := 7 12 14 12; display {(i,12) in B}; # said ":= 14" instead of ":= 7 14" Fix bugs in "expand" and "solexpand" of a dropped objective or constraint. Example: param N := 5; var x{1..N} >= 0; fix12: x[1] + x[2] <= 0; convex: sum{i in 1..N} x[i] = 1; minimize zot: sum{i in 1..N} i*x[i]; blotch: sum{i in 1..N} x[i] + sum{i in 1..N} x[i] = 2; fix_x4: x[4] = .3; drop blotch; expand blotch; # memory fault solexpand blotch; 19960116 Fix bug in handling some file names of the form (i), as in for{i in S} commands (i); 19960125 Fix a bug that could cause the error message "invalid refct 0 in opgen" under conditions that seem hard to describe. 19960202 Adjust DOS .exe files to print nonzero return codes from solve and shell commands. Fix bug in handling a defined variable after some change causes a nontrivial linear, piecewise-linear, or nonlinear part to go away. Example: var x := 2; param t default 1; var y = t*(x + 1)^2; minimize zot: x^2 + y^2; print zot; # 85 let t := 0; print zot; # should be 4; was 85 again 19960207 Fix a bug revealed by an example with two objectives, each involving variables with the same literal subscripts, in a loop that alternated between objectives and forced them to be regenerated. A symptom of the bug was that a linear objective suddenly appeared nonlinear. Example: param p; var x{1..2, 1..2} >= 0; minimize zip: p*x[1,2] + 2*x[2,1]; maximize zap: p*x[1,2] + 3*x[2,1]; s.t. convex: x[1,2] + x[2,1] == 1; option solver cplex; for{i in 1..3} { let p := i; solve zip; solve zap; } Fix botch in handling named problems that could cause unpredictable results (under conditions hard to describe). 19960219 Fix bug (apparently introduced in version 19951019) in handling files that do not end with a newline: the last partial line was ignored, and, under -t, a memory fault was possible. [In the MS-DOS student binaries, this is fixed in version 19960219a.] Fix memory fault in var x <= 3; minimize zot: (x - 4)^2; let x := x.ub; # before something made presolve run 19960223 Fix a bug in handling several slices over the same set. The bug was revealed by a memory fault in a very complicated example. 19960227 Fix memory fault revealed in a "repeat" loop as the first command executed, such as param i default 0; repeat { display i; let i := i + 1; } while i < 3; Fix glitch in exiting "for" and "repeat" loops after an error: for example, on infeasible problems, the sequence "expand; solve;" caused enclosing loops to terminate after the "expand" without for loops being torn down properly; this could make dummy variables of the "for" loops still appear to be known. Example: var x >= 0; param p; maximize foo: x; s.t. zot: x <= p; let p := 4; repeat { for{k in 1..4} { let p := p - 1; display p,k; expand; solve; display x; } } while p > -3; Fix bugs in, e.g., "display _svar, _svar.ub, _svarname;" and in evaluating _svarname[i]. 19960306 Allow printf's "+" flag to reveal the sign of 0. Fix glitch (e.g., memory fault) with redeclaring variables, constraints, or objectives when a subsequent "solve", or "solution" command occurs before anything else forces the problem to be reinstantiated (as the redeclaration should have been doing). Fix bug with "option presolve 0" when bounds turn piecewise-linear terms into linear terms. 19960313 Fix bug in handling index sets common to several entities when some change (e.g., a let command) caused some of the entities to be recomputed. It was possible for the common index set to be deleted prematurely, leading to subsequent chaos. Arrange for redeclarations of variables, constraints, and objectives to retain current values and drop/fix status. Fix bug introduced 19960223 in handling multiple occurrences of the same slice of a subscripted set. 19960330 Fix glitch (possible memory fault, or wrong subscripts in error messages) with dummy variables instantiated in linear sums appearing in objectives and constraints. Fix glitch in evaluating logical expressions: in (A and B) with A false or (A1 or B) with A1 true, if B appeared several times, it might be evaluated, possibly leading to an error message. Example: set I := 1..2; param p{i in I, j in I: j >= i} := i + j; set S := {i in I, j in I: j >= i && p[i,j] > 1 && p[i,j] < 3}; var x{S}; minimize zot: sum{(i,j) in S} (x[i,j] - p[i,j])^2; solve; # "invalid subscript p[2,1]" New call command for directly invoking user-defined functions for their side effects: rather than, e.g., saying "print foo(1,2,3);" or "let Dummy := foo(1,2,3);" to force foo(1,2,3) to be called, you can now say "call foo(1,2,3);". Syntax: call [indexing] function [(arglist)]; 19960403 Fix glitches introduced 19960401 during "improvements" in set generation. For example, sets with "such-that" conditions that do not depend on anything in the set were effectively ignored. Example: display sum {i in 1..2: 0 == 1} i; should print 0 (but printed 3 with the bug). 19960411 Banish "adjoin: SBUFLEN is too small" message arising from reading a large "param foo := ..." table in a data section. Fix bug in handling updates to sets that turn out to be the same. Example: set A; set B; let A := {'abc'}; let B := A; for{i in 1..2} { let A := A union {i}; let B := B union {i+10}; } display A, B; # said no data for set A # listed B twice with the "for" changed to "for{i in 1..1}" Fix memory fault with nested "for" commands in which an inner "for" has a "such that" clause that causes the inner "for"'s body not to be executed on the last iteration of the outer "for". Example: param p; for{i in 1..3} { let p := 10 + i; for{j in 11..13: j > p} print i,j; } 19960417 Fix bug (memory fault) in handling constant subscripts in problem declarations. Fix bug in recording partial drop statuses in certain circumstances; the bug could cause the erroneous conclusion that all members of an indexed collection of variables were dropped. Example: var x{1..4} >= 0; var y{1..4} >= 0; minimize foo{i in 1..2}: sum{j in 2*i-1..2*i} j*x[j]; minimize goo{i in 1..2}: sum{j in 2*i-1..2*i} j*y[j]; s.t. zip{i in 1..2}: sum{j in 2*i-1..2*i} x[j] = 1; s.t. zap{i in 1..2}: sum{j in 2*i-1..2*i} y[j] = 1; problem p{i in 1..2}: x[2*i-1], x[2*i], foo[i], zip[i]; problem q{i in 1..2}: y[2*i-1], y[2*i], goo[i], zap[i]; solve p[1]; solve q[1]; # erroneously said "all variables fixed" Following Fortran 90, add "/=" as a synonym for "!=". 19960418 Fix bug introduced 19960331 with read commands with "<" redirections. Terminate includes within the file in a read command's "<" redirection when the read command ends. 19960425 Fix buglet with the builtin "num" function: it complained at trailing white space rather than ignoring it. Fix bug (possible memory fault) with displaying strings involving '$' followed by a numeral. Example: display 'ab$123'; # memory fault Fix bug printing an error message for a symbolic parameter that fails an assertion (such as an "in" clause) in its declaration (after being assigned a value by the "read"command). Example: set S := {'A', 'BCD'}; param p symbolic in S; read p; E display p; # fault Fix bug (memory fault, or worse) with "purge" and "delete" commands applied to problems. Fix erroneous message about circular dependencies in redeclarations of arcs. Fix memory fault with certain sequences of redeclarations and expression evaluations involving variables fixed by presolve. Example: var x; s.t. f: x = 0; display x.lb0; redeclare var x >= 0; display x.lb0; # fault Adjust xref command so it does not show nodes cited in arc declarations. Make it possible to delete and purge arcs. Fix bug (e.g., memory fault) freeing "continue" commands after execution of compound commands containing them. 19960510 Fix bug with the display command's handling of unsubscripted _con references: chaos could sometimes ensue if there were more constraints than variables (with numbers on different sides of a power of 2). 19960516 Fix bug with "reset data foo;" when foo is not declared with an indexing set. Unpredictable results were possible. Fix bug displaying symbolic parameters that look like numbers. Example: param p symbolic := '1D23'; display p; # said p = '1e23' 19960516 Fix bug with "reset data foo;" when foo is not declared with an indexing set. Unpredictable results were possible. Fix bug displaying symbolic parameters that look like numbers. Example: param p symbolic := '1D23'; display p; # said p = '1e23' 19960606 Fix glitch with "cd" command, which could fail if it involved a param that had not been evaluated previously. Adjust option debug_stub to honor (and retain) ($solver)_auxfiles. Fix bug in handling loops containing data commands: more than a dozen or so iterations elicited an erroneous message about includes being nested too deeply. Example: model diet.mod; for{1..20}{update data; data diet.dat;} Fix bug in handling nonconvex (resp. nonconcave) piecewise-linear terms: when the problem had "unused" variables, either because some indexed variable had an indexing set larger than the set of indices used in the problem, or because some previous command forced another variable to be instantiated that did not appear in the current problem and was declared before one of the problem's variables, the resulting ".nl" file had an incorrect first line. Example: var x{i in 0..2} >= -20; # x[0] is "unused" convex: x[1] + x[2] == 1; minimize z: <<-1,1,4; -1,1,-1,1>> x[1] + <<4;-1,1>>x[2]; Fix some single-stepping glitches, e.g., in handling a command entered in single-step mode that fails for want of data. Fix nasty but seldom-seen bug in computing the right-hand side of a set-valued "let" command and the indexing set of a "for" loop (but not other indexing sets): if the set was represented by a memory previously used for an indexing expression whose scope involved common expressions that changed with the set members, the set was treated as empty. Fix some memory leaks (in scripts that execute many commands). Fix glitch with "continue foo;" in a "for" loop within loop foo: dummy variables of the inner "for" were not restored properly, which could cause subsequent parsing errors in declarations or commands involving the dummy names. Pretend a semicolon appears at the end of command files that end with a compound command with optional final parts missing: repeat ... { ... } # no final condition or semicolon if ... then { ... } # no else clause 19960609 Fix bug in computing recursively defined params when several independent recursions are needed. Example: set A circular := 1..146; set B circular by A ; param c { d in A } in B := if d in B then d else c[next(d)] ; data; set B := 1 4 21 26 36 39 44 57 60 68 71 94 97 101 134 137 142 146; display c; #fault Diagnose (illegal) circular recursive set definitions (rather than looping until memory is exhausted). 19960615 Fix bug in handling "repeat" commands with a "while" or "until". Under complicated conditions, a fault was possible. Fix glitch with "expand" and "problem": an "expand" command executed as the first command after various declarations, including problem declarations, sometimes resulted in a surprising error message ("unexpected type ... in collect_setup()"). Fix bug in the interaction of named problems and drop/restore, fix/unfix: trouble was possible if explicit "drop" or "fix" commands eliminated all members or "restore" or "unfix" commands brought them all back into the problem. Fix bug (memory fault) with the sequence s.t. foo{...}: ...; #new constraint declaration display foo; solve; # memory fault after the solver message is echoed. Fix bug, probably introduced in version 19960330, in handling the logical expressions not (a == b) and not (a != b) when a or b was symbolic (e.g., a symbolic param). Fix glitches with option single_step: there was no break at the first command of a compound command, and using control-D (Unix) or control-Z (MS-DOS) to single step the last command of a compound command resulted in EOF. 19960627 Fix bug simplifying piecewise-linear expressions: example: var x; minimize foo: <<1;-1,1>> x + <<2;3,4>> x; was turned into <<1, 0; 2, 4, 5>>x rather than into <<1, 2; 2, 4, 5>x, but var x; minimize foo: <<2;3,4>> x + <<1;-1,1>> x; was handled correctly: there was a glitch in one of the end cases. Fix glitch with cd command in Symantec DOS AMPL.EXE: explicit letters were interpreted off by one. New option (for debugging nonlinear models): option nl_permute (default 3) tells whether to permute constraints and variables as described in "Hooking Your Solver to AMPL". The value, mod 4, tells what to permute: 0 nothing 1 just constraints 2 just variables 3 both constraints and variables Note that some solvers, such as cplex, minos, and osl, require the permutations. 19960707 Fix glitch with var = declarations with "constant" right-hand sides (not involving any variables, i.e., declarations that should have said "param" rather than "var"): in some cases, such variables did not appear to have been initialized with the right-hand side value. Example: var x = 3; var y = 4; display x; # correct: x = 3 display y; # wrong: y = 0 Fix bug in updating slices over sets declared without := or default. Example: set A; set B dimen 2; param p{i in A, (i,j) in B} := i + j; display B; # provokes error message: no data for B set C; data; set C := a b c; set A := 1 2 3, 10, 20, 30, 40, 50; set B := (1,10), (1,20), (2,30), (3,40), (3,50); display p; # fault Quietly omit defined variables from problem declarations. (This repairs a memory fault previously associated with such declarations.) 19960709 Prevent execution of problem commands from running presolve (and thus possibly giving an erroneous message about inconsistent constraints). 19960718 Fix bug with "option nl_permute 0" and "option nl_permute 1": wrong information on Jacobian nonzeros sometimes went into the .nl file. Fix bug in reading .sol files corresponding to option nl_permute at nondefault values. 19960729 Fix glitch in reading data sections having sets of arity > 10. Disallow writing MPS files for problems with nonlinear variables. Fix glitch in reading g-format .sol files written with \r\n at the ends of lines. Omit INITIAL bounds from .mps files; they were only needed long ago by minos. New builtin functions: time() returns the current time (in seconds since 00:00:00 1 Jan. 1970 GMT); ctime() returns a 24-character string representing the current time, and ctime(t) returns a similar time representing time t (or the current time, if t is not numeric). 19960731 Fix bug in numbering variables when an error (such as "no value" for a param) interrupts problem instantiation. Symptom: wrong variable names in "expand" command. The bug bit under complicated conditions. 19960816 Fix glitch in the numbers of nonlinear and linear variables reported by option show_stats when option nl_permute has the nondefault values 0 or 1. Fix faults under complicated conditions (a) after discarding invalid subscripts and (b) with missing data. Fix bug in the expand command applied to a constraint or objective declared with "{if test}" when "test" turns out false, as in param a := 1; var x; s.t. c{if a > 2}: x >= 1; Fix infinite loop in processing the erroneous input var x; s.t. con2: {if 1 > 2} x >= 1; # syntax error (misplaced : ) expand con2; Fix bug in naming constraints and objectives when an {if false} constructs appear. Example: var x; s.t. c {if 1 > 2}: x >= 1; expand c; s.t. con2 {if 1 <= 2}: x >= 1.5; expand con2; # con2 >= 1.5; Fix bug (fault) in "delete" applied to a pipe function. New predefined param _pid gives the process ID of the AMPL process (a number unique among processes running on the system). New option format_range_errmsg, if not '', replaces %i, %d, %o, %u, %x, or %X in printf commands or sprintf invocations in which the number to be converted is outside the appropriate range of integers. If $format_range_errmsg has its default value of '', an error message appears and the command is aborted. New automatically determined set: _AVAILFUNCS = user-defined functions that have been linked (statically or dynamically) with AMPL and thus can be declared and evaluated in the AMPL session. (Other user-defined functions may be declared and used in constraints and objectives, but AMPL will not be able to evaluate them. It can, however, pass them by name to solvers that know how to evaluate them.) New automatically updated params give some problem statistics the solver sees: _snbvars = number of binary variables used linearly _snivars = number of general integer variables used linearly _snnlcons = number of nonlinear constraints _snnlobjs = number of nonlinear objectives _snzcons = number of nonzeros in the constraint Jacobian matrix _snzobjs = number of nonzeros in objective gradients 19960819 Fix glitch in handling "default nnn" in a data section when not followed by ":=" and when applied to a param indexed by a set of arity > 1. Example: set A dimen 2; param p{A}; data; set A := a b c d e f; param p default 3 [c,d] 4; # default phrase appeared display{(i,j) in A} p[i,j]; # to be ignored. 19960822 Fix a glitch that sometimes caused a "bad binary file" message from the "solve" and "solution" commands on the DEC Alpha. The glitch was in handling 64-bit "long" integer variables and had no effect on most (32-bit) machines. 19960825 Fix a bug in handling common set expressions whose first uses involve membership tests and whose later uses involve iteration over the set expressions, as in set A; set B; ... s.t. c1{i in C}: ... if (i in A union B) then ...; s.t. c2: ... sum{i in A union B} ...; solve; # Change something that will force c1 and c2 to be regenerated solve; # possible fault or worse 19960826 Fix botch in error message given when a data section provides a subscript for a set that is declared not to be subscripted. The botch could cause a fault or a strange name in the error message. Example: set A; data; set A[1] := a b c; 19960829 Add "inout" and "out" to the reserved-word list (for use in upcoming extensions). 19960922 Fix glitch with option log_file under block mode: some input didn't get copied to the log file. Fix performance bug (which slowed execution and burned memory, but did not affect correctness of results) in handling slices of subscripted sets. Terminate all (not just some) commands in which genmod detects errors. 19961003 Fix bug (leading to a memory fault during execution) in parsing superfluous braces within a compound command, such as those around the display command in for{i in 1..2}{ if i < 2 then print 'yes'; {display i;}} 19961016 Fix glitch (loop) in recovering from a syntax error in a list of commands gratuitously enclosed in braces. Fix bug that bit (e.g., with a fault) under complicated conditions with recursive params (including those with a recursive "check" condition) that have a default expression. Fix bug with option linelim: in a model with defined variables, the sequence option linelim 1; solve; /* adjust something, then... */ option linelim 0; solve; display _var; could lead to a fault (or worse). 19961017 Fix fault with defined variables in certain sequences of commands, such as display a_defined_variable; problem another_problem; let variable := ...; New option show_write_files controls whether the write command reports the names of the files it writes: 0 ==> no (default) 1 ==> yes for "write;", no for "write gstub;" 2 ==> yes 19961023 Fix bugs with ord(i), next(i), prev(i) and their relatives when i is a dummy in a for loop. 19961100 Fix some bugs in handling subscripted environments and problems: faults were possible under the right complicated conditions. Example: environ E{1..3}; set I := 1..2; var x{i in I} >= 0; minimize zot: sum{i in I} i*x[i]; s.t. convex: sum{i in I} x[i] = 1; problem foo{i in 1..3} environ E[i]: x, zot, convex; problem foo[1]; # fault Fix bug in "show e;" (to show environment names) after a "reset;" following environment declarations (including the implicit ones associated with problem declarations). 19961109 Fix bug with defined variables: under certain conditions, a scalar or subscripted defined variable evaluated as 0 rather than its correctly computed value. (The situation in which this happened could be seen with "option times 1;": the bug bit when collect ran without a subsequent presolve before the defined-variable evaluation.) 19961112 Fix glitch with "redeclare": it faulted when applied to a declaration for a name that had not been declared (or had been deleted). 19961130 Fix glitch with recursive set declarations: in situations where the same expression appears twice or more, it was sometimes evaluated when not needed, possibly giving rise to an erroneous error message about an invalid subscript, as in the following example: param p{i in 2..4} := i; set S{i in 1..4} := if i == 1 then {} else if p[i] == 2 then {1} else if p[i] == 3 then 2..3 else 1..4; display S[1]; # error message about evaluating p[1] Fix fault with evaluating certain nonlinear "if" expressions in which the condition does not involve variables and can be evaluated in an outer context. The show command's printing of recursive parameter definitions will now say ":= /*recpd.*/(...)" or ":= /*srecpd.*/(...)" rather than ":= recpd.(...)" or ":= srecpd.(...)". 19961210 Adjust most commands (but not declarations) to permit unqualified objective, constraint, and variable names in () expressions. Hitherto one had to use objective.val, constraint.dual, and variable.val. Now one can write something like option cplex_options (sprintf('... uppercutoff=%.f ',_obj[1])); (rather than _obj[1].val) to adjust an option to reflect an objective value from the most recent solve. Disallow expressions involving variables in arguments to member, ord, ord0, next, nextw, prev, and prevw, when they appear in constraint and objective declarations. 19970104 Fix another bug in recursive set definitions involving literal sets (such as {1}), setof expressions and iterated unions. Unpredictable results were possible. 19970212 Fix rarely seen bug (resulting in a memory fault) that arises under complicated conditions. 19970219 Fix (rarely seen) bug that could cause a fatal "solve_out bug!" message. Work-around: "option presolve 1;". Fix failure to diagnose nonnumeric values in the following context: param p{1..2}; data; param p := 1 abc 2 xyz; display p; # bogus values for p Fix bug (memory fault) in deleting defined variables. In data sections, recognize as numeric values Infinity, -Infinity, +Infinity and, on IEEE-arithmetic systems, NaN, -Nan, +NaN. Code around WATCOM bug in comparing NaNs: the bug caused the display command to turn NaNs into zeros. 19970307 Complain at the appearance of problem names in the itemlist of a problem declaration. Tweak "show problem_name;" to print the itemlist in the declared order (rather than its reverse). Arrange for variable values of 0 assigned by "let" to be explicitly conveyed to solvers unless $reset_initial_guesses is nonzero. Previously they were explicitly conveyed only after a solve, or if they were given in a data section or variable declaration. 19970314 Fix bug in evaluating expressions involving problem-dependent synonyms (such as _nvars and _snvars): sometimes, e.g., when they appeared in printing commands after a problem command had changed the current problem (and before anything else caused the new problem to be instantiated), they were not recomputed. Fix memory leak with nonlinear expressions (in constraints and objectives). Arrange for "option gentimes 2" to show total memory rather than incremental memory. Tweak -T output so "genmod times" lines (other than headers) always start with "## ". This was hitherto only true for seq numbers <= 99. 19970402 In the (weird) context set I := 1..2; var x{I}; for{i in I} if i >= 2 then print i,3; minimize zot: sum{i in I} i*x[i]; before processing the "minimize zot" declaration, recognize that there is no "else" command to match the "if" command. When a program (e.g., solver) cannot be invoked (under Unix), say why. Adjust generation of file names with option debug_stub to avoid confusion when all variables are fixed by presolve. Fix glitch (possible fault) introduced 19970212 affecting string functions ctime, gsub, sprintf, sub, substr. 19970408 Fix glitch in handling redirections in the "read" command: entities in them were not instantiated by the read command. Example: param dummy; param fname symbolic := "seq.dat"; read dummy < (fname); # error message: fname not instantiated display dummy; Adjust the handling of the default-value symbol in data sections so it is only recognized when given with the same kind of quotes (none, " or ') with which it is declared. (The initial default-value symbol continues to be . with no quotes.) Thus, for example, param p symbolic; data; param p := "."; display p; will now show that p has the value "." rather than no value. 19970413 Under MSDOS (and MSWIN), quietly ignore command-line options -v2 and -ve (which are always in effect). Fix bug in handling "if" expressions in some contexts when the "if" test can be done at a more outer loop level than the "then" or "else" expressions. Example: param j := 1; set T := 1..10; param x{T}; let{t in T} x[t] := max(0, if j == 0 then t else 0.5*t); display x; # all x components had the x[1]'s value (0.5). 19970417 Fix bug that could give "tva top error" message when evaluating complicated recursive parameters or sets (when first evaluated in certain contexts, such as a printing command). Example: param M default 5; set I := 0 .. M; set A := {i in I, j in I, k in 0..1: (i < M || k == 1) && (j < M || k == 0)}; param N{(i,j,k) in A} := if max(i,j,k) == 0 then 1 else 1 + (if k == 1 then (if i < M then N[i,j,0] else N[M-1,j,1]) else if i > 0 then (if j == M then N[i-1,M,0] else N[i-1,j,1]) else N[M,j-1,1]); display N; Workaround: declare something that forces the recursive entity to be evaluated. For example, changing the display command to param Nlast := N[M-1,M,0]; display Nlast, N; bypasses this bug. (Uses of recursively defined entities within a model bypass the bug.) Suppress (e.g.) printing commands when they involve a wrong number of subscripts. 19970420 Fix bug introduced 19970402 in handling the trailing opt_while of a repeat command. 19970516 Fix trivial performance bug in display. Fix glitch that could cause an erroneous missing data message in a command immediately following an error message (under the right conditions). Fix infinite loop with a nested commands 'foo'; command when foo has the form if something then {...}; (with no outermost else). Fix bug with indexed collections of problems: under just the "right" conditions, the actual indexing set of the problems was aliased with other sets, wreaking havoc when a new member was added to the actual indexing set (if the aliased sets should remain fixed). Fix bug with "redeclare": dependents shown by "xref" were not updated properly. 19970517 Fix bug in delete, purge, and redeclare commands: sometimes an internal structure was deleted twice, leading to chaos. Example: model nlmodels/ljcluster.mod; s.t. constr {i in I, j in I: i < j}: r[i,j] >= 1e-6; solve; # error in constr[1,2]: bad subscript redeclare s.t. constr{(i,j) in P}: r[i,j] >= 1e-6; solve; # fault Fix glitch in show command: default (initial) values for variables were not printed. 19970528 Fix glitch in comparing dummy indices and symbolic parameters with character strings that look like quoted numbers: the comparison proceeded as though the quotes were not there. Example: set A; data; set A := a 2 c; display {i in A: i != '2'}; gave set {i in A: i != '2'} := a c; (as though the quotes around '2' were dropped) rather than (the correct) set {i in A: i != '2'} := a 2 c; Fix bug in _sobj[...]: objectives presented to solvers are permuted so nonlinear objectives come first; _sobj[...] did not reflect this. Example: set I := 1..2; var x{i in I} >= 0 := i; minimize zip: sum{i in I} i*x[i]; minimize zap: sum{i in I} i*x[i]^2; s.t. convex: sum{i in I} x[i] = 1; print _sobj[1], _sobj[2]; gave "5 9" rather than "9 5". Extension to expand command: recognize synonyms (possibly subscripted _con, _obj, _var, _scon, _sobj, _svar). 19970529 Fix failure to recompute reduced costs after a relevant change to $abs_boundtol or $rel_boundtol. 19970606 Fix a bug with sequences of named problems of increasing sizes: it was sometimes possible for a memory block to be put on the wrong free-list, causing unpredictable subsequent havoc. 19970609 Fix glitch with "check;" in single-step mode. It elicited a mysterious (but harmless) message about "bad type 73 in cmdstring". 19970615 Fix fault that could arise, e.g., if an option command is the very first command to be executed and it follows a problem declaration. Adjust the tests involving $pl_bigM (done when linearizing nonconvex piecewise-linear terms) to assume a lower bound of (smallest breakpoint) - $pl_bigM and upper bound of (biggest breakpoint) + $pl_bigM on the variable "multiplied" by a piecewise-linear coefficient when the true bounds on the variable exceed these values (instead of just when the true bounds are infinite). 19970721 Change "arity" to "dimen" in some error messages. Fix bug in writing .env files (with "solve" or "write" commands when $solver_auxfiles or $auxfiles contains 'e') and shell commands (when $shell_env_file is not empty): names were listed twice. Fix bug in generating indexed sets for some declared := right-hand sides that do not depend on the set's index. At least two surprising error messages were possible, depending on the right-hand side. Examples: set A; set B{1..2} := A; data; set A := a b c; display B; gave error processing set A: no data for set A and set S dimen 2; set T := setof{(i,j) in S} i; set U{T} := setof{(i,j) in S} j; ## probably should be set U{i in T} := setof{(i,j) in S} j; data; set S := a b a c b c; display U; gave set::rehash finds duplicate entry 19970729 Fix bug with random functions that caused some expressions involving them not to be reevaluated. Example: set S default {}; repeat { let S := S union {Uniform(0,15)}; } while(card(S)) < 5; looped infinitely. Change exit to a command of the form exit ; or exit expression ; The expression, if given, must be numeric and specifies the process return code; an omitted expression is treated as 0. Change "quit;" to be equivalent to "exit ;", which means it may now appear in compound commands. Here, stands for the return code that "quit" has always given, which reflects recent syntax errors (and is cleared to 0 by successful command executions). Fix some bugs with "break commands;" and "break all;"; adjust these commands so if no "commands" command is being executed, they terminate the current nest of includes, if any, and otherwise are no-ops. 19970731 Fix bug in presolve for problems with integer (or binary) variables: for $presolve > 1, if a constraint with two remaining nonzeros caused an integer variable to be fixed and the constraint had previously implied a bound on the other variable, that bound was not conveyed to the solver. (Note that $presolve is 10 by default; specifying "option presolve 1;" avoids the bug.) Add a Caution for writing an MPS file for problems with integer variables having infinite bounds. There's no "standard" MPS way to express such problems. 19970814 "Invisible" tweak to presolve changes of 19970731. With invocations with -o (and no commands that would cause the -o command-line option to be ignored), if presolve fixes all variables, say so and give process return code 16 rather than 0. New command-line option -R (recognized only as the first command-line option and not mentioned in the "-?" listing of options) puts AMPL into "server mode", in which it declines to execute cd and shell commands, forbids changes to options TMPDIR, ampl_include, and PATH (or whatever is the usual name of the search path for the operating system being used), disallows pipe functions, and restricts names in option solver and file redirections to be alphanumeric (so they can only write to the current directory, which, on Unix systems at least, cannot be changed). By invoking ampl from a shell script that suitably adjusts the current directory and environment variables TMPDIR, ampl_include, and PATH before it invokes "ampl -R", one can control the environment in which "ampl -R" operates. 19970820 Fix bug in renumbering variables with certain sequences of commands that do not cause presolve to run (it runs in some form regardless of option presolve) and do cause variables to be instantiated "out of order". On Unix systems, adjust printing of commands with -T and -t to appear on stderr rather than stdout. Omit variables and constraints added to linearize piecewise-linear terms from the synonyms (_var, _con, etc.). They sometimes caused faults with "display _con.slack;" and "display _var.rc;". 19970912 Fix bug with scalar defined variables: linear terms were not merged. Fix possible glitch (usually invisible) in distinguishing network constraints from linear constraints: constraints added to linearize piecewise-linear terms may sometimes have been misclassified. 19970922 Fix bug with the let command: assignments to an indexed symbolic parameter for a subscript for which the parameter previously had no value could fault. Under -R, forbid changing shells. 19970925 Fix glitches affecting handling of some complementarity constraints (an extension to be described later). 19971003 Fix bug with "redeclare ..." of entities on which other entities depend: the other entities were sometimes not recomputed when the redeclared entity changed. Fix a memory leak: in loops (or scripts) where indexing expressions of declared entities changed values, memory for the indexing set was not reclaimed. 19971006 Fix fault with "set unordered_set := ordered_set;"; example: set A ordered; set B := A; data; set A := a b c; display B; # fault 19971110 Fix bug with symbolic-valued "if" expressions whose test could be lifted to an outer loop: on Sun machines (because of Sun's calling conventions) and possibly others (with more complex examples), the expressions appeared to be miscomputed (and could lead to a subsequent fault). 19971117 Fix glitch in .nl headers for Microsoft systems that probably only matters for complementarity solvers: the eqns number was omitted. 19971205 Fix a bug with recursive parameters involving certain expression lists, such as those in min and max expressions. It may be possible to construct an example where the bug led to an incorrectly computed result, but this does not seem easy to do. The bug did cause a fault on one machine (with a complicated example), but was invisible on several other platforms. Fix glitch with objectives and constraints involving imported functions with symbolic (string) arguments, which elicited the surprising error message "unexpected type 0x13c8 in simplify()". Fix bug with "write 0;" that could cause a fault or other untoward behavior with subsequent solve and write commands. 19980102 Fix presolve bug: if a nonlinear constraint was eliminated but had two or more unfixed variables at the time the last nonlinear variable was fixed, the nonlinear expressions were ignored. Example: var x{1..4}; s.t. c1: x[1] = 2; s.t. c2: x[2] = x[1] + 1; s.t. c3: x[3] = x[1] + 2*x[2]; s.t. c4: x[4] = x[1]^2 + x[2] + x[3]; solve; display x, _con.slack; # x[4] = 11 rather than 15 Fix an infinite loop (or worse) that could be encountered with a command-line invocation to write a .nl file, as in ampl -obfoo foo.mod foo.dat (with no commands in the foo.* files). The bug bit when a common expression appeared several times in certain contexts. For example, "0" is such an expression in the following (silly) example: param p in [0,10) default 3; param q in [0,10) default 4; param r in [0,10) default 5; var x := p*Uniform(1,2) + q*r; minimize zot: (x-6)^2; 19980127 Fix bug in handling variable.dual for fixed variables. Note that variable.dual is 0 unless it's a defined variable, in which case variable.dual is the dual value for the defining constraint; defined variables cannot be fixed. Usually variable.rc, the reduced cost, is of greater interest: it's the dual value for bounds on the variable. New syntax for expressing complementarity constraints: in addition to previous forms, constraint declarations may now have the form constraint_start : constr1 complements constr2 ; in which constr1 and constr2 consist of 1, 2, or 3 expressions separated by the operators <=, >=, or == (or =). As with other constraint declarations, constraint_start gives a name to the constraint, an optional descriptive string (alias), and, if the declaration describes a collection of constraints, an indexing expression. In constr1 and constr2 together, there must be a total of two explicit inequality operators, with == counting as two. A complementarity constraint is satisfied if both constr1 and constr2 hold and at least one inequality is tight, i.e., satisfied as an equality. If one of constr1 or constr2 involves two inequalities, then the constraint must have one of the forms expr1 <= expr2 <= expr3 complements expr4 expr3 >= expr2 >= expr1 complements expr4 expr4 complements expr1 <= expr2 <= expr3 expr4 complements expr3 >= expr2 >= expr1 In all of these cases, the constraint requires the inequalities to hold, with expr4 >= 0 if expr1 == expr2 expr4 <= 0 if expr2 == expr3 expr4 == 0 if expr1 < expr2 < expr3 For expressing MPECs (math. programs with equilibrium constraints), complementarity constraints may coexist with other constraints and objectives. The suffix notations, such as constraint.lb, constraint.body, etc., are extended so that for a complementarity constraint, constraint.Lsuffix and constraint.Rsuffix correspond to constr1.suffix and constr2.suffix, respectively, and complementarity_constraint.slack (or the unadorned name) stands for a measure of the extent to which the complementarity constraint is satisfied: if constr1 and constr2 each involves one inequality, then the new measure is min(constr1.slack, constr2.slack) (which is positive if both are strictly satisfied, 0 if the complementarity constraint is satisfied exactly, and negative if at least one of constr1 or constr2 is violated). For constraints of the form expr1 <= expr2 <= expr3 complements expr4, the .slack value is min(expr1-expr2, expr4) if expr1 >= expr2 min(expr2-expr3, -expr4) if expr3 <= expr2 -abs(expr4) if expr1 < expr2 < expr3 so in all cases, the .slack value is 0 if the complementarity constraint holds exactly and is negative if one of the requisite inequalities is violated. Solvers see complementarity constraints in the standard form expr complements lb <= variable <= ub A new synonym, _scvar{i in 1.._sncons}, indicates which variable, if any, complements constraint _scon[i]: if _scvar[i] in 1.._snvars, then variable _svar[_scvar[i]] complements constraint _scon[i]; otherwise _scvar[i] == 0, and _con[i] is an ordinary constraint. Other new synonyms: _cconname{1.._nccons} are the names of the complementarity constraints as the modeler sees them. A forthcoming paper ("Expressing Complementarity Problems in an Algebraic Modeling Language and Communicating Them to Solvers", by Michael C. Ferris, Robert Fourer, and David M. Gay) discusses the new complementarity syntax. The paper will be available from the AMPL web site (http://www.ampl.com/ampl/) after it has gone through the Bell Labs release process. Anticipating extensions to be described later, the following words IN INOUT LOCAL OUT suffix are now reserved. This breaks the blend.mod and blend.dat distributed with the AMPL book, in which it's necessary to change OUT to Out and IN to In. 19980128 Introduce some new automatically updated params that give information about the current problem, either as seen by the modeler (_nccons), or by the solver (names that start with "_s"): Name Meaning _nccons No. of complementarity constraints before presolve _snbvars No. of binary (0,1) variables _snccons No. of complementarity constraints after presolve _snivars No. of general integer variables (excluding binaries) _snlcc No. of linear complementarity constraints _snlnc No. of linear network constraints _snnlcc No. of nonlinear compl. constrs.: _snccons = _snlcc + _snnlcc _snnlcons No. of nonlinear constraints _snnlnc No. of nonlinear network constraints _snnlobjs No. of nonlinear objectives _snnlv No. of nonlinear variables _snzcons No. of constraint Jacobian matrix nonzeros _snzobjs No. of objective gradient nonzeros 19980202 Fix glitch in user-defined suffixes (to be described later). 19980208 Fix a bug with card(), first(), and last() that occasionally caused unpredictable behavior. Fix bugs (leading to invalid .nl files) with "option linelim 1" on some problems with nonlinear defined variables. New builtin suffixes: variable.defeqn is the subscript of _con for the corresponding defining constraint if the variable is a defined variable, and is 0 otherwise. Similarly, constraint.defvar is the subscript of _var for the resulting defined variable if the constraint defines a defined variable (either through the var = syntax or because of option substout), and is 0 otherwise. New options log_model and log_data (default 0): if option log_file is not '', declarations and commands in included files are copied to $log_file if $log_model is 1, and included data sections are copied to $log_file if $log_data is 1. Edit change description of 19950619 above to describe _display. For convenience, here is a summary of _display and csvdisplay. Commands _display and csvdisplay are variants of display that emit tables in a more regular format than does display: each line of a table starts with s subscripts and ends with k items, all separated by commas. _display and csvdisplay differ in the table headers they emit. The header for _display consists of a line starting with "_display" and followed by three integers s, k, and n (the number of table lines that follow the header), each preceded by a space. Whether csvdisplay emits a header is determined by option csvdisplay_header (default 1): if $csvdisplay_header is 1, cvsdisplay's first line has the form Index_1,Index_2,...,Index_k,Expr_1,...,Expr_n (where k is the number of subscripts and n the number of expressions in one line of the table that follows); if the jth expression is a simple name (not of the form Index_i or Expr_i) that has not appeared previously in the current csvdisplay header line, then that name appears rather than "Expr_j". If $csvdisplay_header is 0, this header line is omitted. Options csvdisplay_precision and csvdisplay_round govern the precision of printing for _display and csvdisplay; the defaults (0 and '') give full precision. 19980212 Fix bug with let subscripted_set[...] := set_expression when subscripted_set[...] had no previous value. If the bug bit, it could cause unpredictable behavior. Fix bugs simplifying nonlinear "if" expressions after presolve fixes a variable: if the result of the test condition was now known, it was sometimes misinterpreted; otherwise, expressions whose evaluation the test condition should prevent sometimes led to error messages, such as a complaint of division by 0 (for a quotient whose value presolve had determined to be 0). Examples: var x{1..3} >= 0; s.t. fix_x1: x[1] = 0; s.t. zot: if x[1] < -1 then 3 / x[1] else x[3] + x[2] == 1; and var x{1..3} >= 0; s.t. fix_x1: x[1] = 0; s.t. zot: if x[2] < -1 then 3 / x[1] else x[3] + x[2] == 1; (Both examples give the same error messages, but the causes differ.) 19980216 Fix bug introduced 19980127 affecting partial drop, restore, fix, and unfix commands executed before the first solve (or other command requiring the affected variable, constraint, or objective to be instantiated). Example (with multi.* from the AMPL book): model multi.mod; data multi.dat; fix Trans; unfix {i in ORIG, j in DEST} Trans[i,j,"bands"]; solve; # fault 19980218 Fix longstanding bug with recomputing sets: an empty set was sometimes not recomputed. Example: for {s in 1..2} display {i in 0..3: i = 5-s}; Fix bug introduced sometime between 19950724 and 19961031 in handling certain recursive parameter and recursive set definitions: under complicated conditions, a fault or wrong value was possible. 19980302 Fix bug in handling sets with default value {}: if two such sets were forced to have their default (empty) values, and then one was assigned to the other, one of them was sometimes later diagnosed to have no data. Example: set A default {}; set B default {}; display A, B; let A := B; display A, B; # complaint about A having no data. Introduce (on all systems other than AIX and MSDOS, but including W95 and NT) a new scheme for acquiring imported functions: execute ampl -i? for more details. More extensive documentation will appear later. 19980305 Fix bugs in handling user-defined suffixes (a new feature to be described later) when certain indexing sets change. Fix memory fault with the delete command in certain sequences of commands and declarations. Example: param n default 3; var x{i in 1..n} default i; minimize silly: sum{i in 1..n} (i - sin(x[i]))^2; param q; solve; param r; let r := silly; delete q; let n := 4; solve; # fault Fix bug introduced sometime between 19950724 and 19961109 in updating some set expressions that depend on a param with a := assignment whose right-hand side depends on another param with a value given in a data section. The first time the latter param was changed, the set was not recomputed. Example: param p; param q := p/2; param r{i in floor(q)..ceil(q) + 1} := i+1; data; param p := 7; display r; # OK let p := 23; display r; # r came out wrong Fix a presolve bug with complementarity constraints. Example: var x; var y; s.t. equ: x + y = 1; s.t. compl: x >= 0 complements y >= 0; s.t. inequ: x + 2*y <= 1; solve; # error message (rather than "solution determined") 19980309 Change to option pl_linearize: the default value (1) now uses a more efficient linearization of nonconvex (resp. nonconcave) piecewise-linear terms and uses suffixes .sos and .sosref to tell solvers that handle SOS sets about the variables and constraints that imply the SOS2 constraints introduced by the linearization. A forthcoming update to the AMPL/solver interface library will provide sample interfaces that exploit this new information. To get the old behavior, specify option pl_linearize 2; Versions 19980305 and 19980308 had glitches in this change. 19980313 Fix bugs with defined variables: in commands, there were miscomputed if nothing else had forced presolve to run, and under "option linelim 1" constant terms were added in twice (in expressions computed in an AMPL session, but not in expressions conveyed to solvers, which saw the correct expressions). Example: var x; var y = x^2 + 1; var z = 3*y + 4; minimize zot: y + z; print z, zot; # wrong values: 0, 0 solve; print z, zot; # still 0, 0 reset data; # clear some internal state solve; print z, zot; # right values: 7, 8 option linelim 1; solve; print z, zot; # wrong values: 10, 11 Fix bug with "option foo;" when foo had the value "." -- quotes around "." were missing. 19980320 In "expand _con[n];", give an error message (rather than faulting) if _con[n] arises from linearizing a nonconvex piecewise-linear expression. It's necessary to use _scon[...] to see these generated constraints. Adjust some details (to be documented later) of importing functions from shared libraries. 19980325 Fix glitch in handling of complementarity constraints with "solexpand;": non-indexed complementarity constraints were sometimes omitted from the "solexpand;" listing. Fix botch in handling complementarity declarations of the form s.t. foo: L <= expr <= U complements v; where v is a variable: the bounds L and U were mishandled. When the solve command cannot open a temporary file, it now tries to erase the file, then open it again (in case permissions prevented opening it the first time). For the benefit of the solver, it also now tries to remove any previously existing temporary .sol file of the name the solver will write. New options shell_exitcode_max (default 2^16) solve_exitcode_max (default 0) cause compound commands to be aborted if a shell command results in $shell_exitcode > $shell_exitcode_max or a solve command results in $solve_exitcode > $solve_exitcode_max . The defaults give the old behavior (except for the unlikely case of shell_exitcode > 2^16). 19980328 Suppress commands involving an invalid name or suffix. Fix glitches with expand _ccon and, for complementarity constraints, expand _con. Warn (rather than faulting) that expand _scvar is not allowed. 19980408 Fix minor inconsistencies with _scconname in constraint names shown by "expand _scon;" and "solexpand cc;", where cc is a complementarity constraint. Warnings of discarded subscripts caused compound commands to terminate but did not inhibit single commands. Now these warnings no longer halt compound commands. Remove prohibition on "reset data" and "update data" commands appearing in compound commands. 19980409 Fix rarely seen presolve bug introduced in version 19980309 that led to an error message of the form "presolve has k = 55, P.nfc = 51". Minor efficiency improvement to some set intersections having subscripted sets as one or both operands. 19980506 Fix bug with delete command: deleting any objective, constraint, or variable other than the most recently declared undeleted objective, constraint, or variable sometimes led to a subsequent fault. Fix bug (that could cause a fault) in testing membership of certain special names in "system sets", such as _PARS, _VARS, _CONS, etc. 19980512 Fix glitch in the handling of tables by cvsdisplay and _display: tables that would be wider than $display_width were not written in the advertised format. Work-around: specify a sufficiently large option display_width. Fix one-bit error in decimal-to-binary conversion of certain denormalized numbers. For example, print 8.44291197326099e-309; gave 8.442911973260987e-309 rather than 8.44291197326099e-309 19980515 Fix glitch in handling imported functions of a variable number of arguments. Sometimes a domain error was erroneously diagnosed. Fix bugs with param p := random_expression; set s := random_set_expression; The random expressions were computed anew at each command that referenced the entities in question. For example, param p := Uniform01(); print p; print p; gave 0.6092090562543065 0.1898730530200304 rather than 0.6092090562543065 0.6092090562543065 19980518 Fix bug with reading large data sections (more than 1024 tokens) that involve "." (or the current default-value symbol) in the first half of the tokens. The default-value symbol was mishandled, possibly leading to a fault or other untoward behavior. 19980520 Fix glitch with the expand command's handling of imported functions. Example: function foo; var x{1..4}; minimize zot: foo({i in 1..3}x[i], x[4]); expand zot; # gave an error message 19980525 Fix glitch in solve_message and solve_result reported when presolve fixes all variables and determines that the problem is infeasible. When "solve;" fails, set solve_result = '?', solve_result_num = -1, and solve_message = '?'. Fix glitches with the display command's handling of builtin dot notations for indexed collections of objectives. 19980529 Fix a glitch that sometimes caused presolve to run unnecessarily, resulting in surprising show_stats output (with option show_stats 1). 19980605 Fix bug in handling _svar.*, _scon.*, or _sobj.* when some entities have been dropped (explicitly or by presolve). Fix bug with fix{...} var[...] := ...; when {...} was affected by the assignment, which happened first: the {...} was recomputed, causing the variables not to be fixed. Recognize _obj.astatus, _sobj.statsus, _obj.astatus_num, and _sobj.astatus_num. The latter two are always 0 (in the problem), the former two "in" unless $astatus_table has been changed from its default value. Fix glitches in "reset data;" and "update data;" that caused previously declared imported functions to be reported as unavailable when invoked. 19980611 Fix bug in computing .astatus values for constraints: the last _nobj (number of objectives) defining constraints (associated with defined variables) were classified as "pre" instead of "sub". Fix bug in sending statuses to solvers when the problem involves piecewise-linear terms. The bug could cause a fault, but was otherwise harmless. Workaround: option send_statuses 0; . 19980615 Fix rarely encountered presolve bug: under unusual circumstances, rounding errors could cause inequality constraints to be erroneously deduced never to be tight. The example that revealed this bug involved an absurd lower bound of -1e38 on some variables. Permit potential defining constraints (other than those introduced by "var ... = ..." declarations) to participate in the strong bound deductions attempted when $presolve > 1, unless $substout mod 4 == 2. Previously, if there were any "var ... = ... " declarations or $substout was nonzero, potential defining constraints were excluded from the strong bound deductions. Specifying "option substout 2" gives the old behavior of "option substout 0", and "option substout 3" gives the old behavior of "option substout 1". The new $substout values 2 and 3 are for possible debugging and their effect may be withdrawn later; values 0 and 1 are the only ones that should normally be of interest. Fix glitch in error messages for failed "within set" clauses in set declarations: "within set" didn't appear in the message (and the attempt to print it might have faulted). Fix bug (probably introduced 19950315) in handling == and != clauses in symbolic parameter declarations: the comparisons were sometimes botched. Adjust logic for checking conditions on sets and parameters: when "update data" and "let" provide new subscripts or modify existing values, just check the new or changed items. This changes the time for some loops from quadratic in the number of iterations to linear. If reports of failed conditions are interrupted by $eexit, the list of questionable subscripts is currently not purged. Otherwise each new value causes only one warning if it fails a condition. Fix a possible memory fault with entities redeclared to be another kind of entity (including symbolic versus nonsymbolic parameters). Example: param p{1..3} symbolic; #... redeclare param p{1..3}; display p; # fault 19980622 Fix glitch in student ampl binaries that could cause a "vT bug" message if one said "solve;" after the message "Sorry, the student edition is limited to 300...". Fix bug in handling next(dummy), prev(dummy), ord(dummy), etc., when the dummy was from a for loop: at the end of the entire compound command, it was sometimes possible for the set over which the dummy ran to up on a free list twice, leading to subsequent unpredictable havoc. 19980625 Fix bug in calling imported functions: the nr (number of real args) field was not set properly. Fix glitch in changes of 19980615 that made _AVAILFUNCS empty (even though imported functions were available). New builtin param _cmdno is 0 initially (or after reset) and is incremented each time a command is executed. New option project_fixed_vars (default 1): if nonzero, presolve projects fixed variables onto their bounds (even if $presolve is 0). If $project_fixed_vars == 2, a message is printed for each fixed variable that is projected (until $presolve_warnings such messages have appeared). It is possible for a variable to appear only in constraints that imply bounds on the variable. When $presolve > 0, such constraints are eliminated. Presolve now projects such variables onto the interval given by their tightest deduced bounds. Here is an example illustrating this and the new option project_fixed_vars: var x >= 3 <= 3 := 4; var y >= 0 := 7; s.t. foo: x + y <= 3.5; display y; fix x; solexpand; display y; Previously, this example gave the output y = 7 presolve, variable y: impossible deduced bounds: lower = 0, upper = -0.5 Infeasible constraints determined by presolve. s.t. foo: <= -0.5; y = 7 The message about infeasibility appeared because "fix x" fixed x at 4, which made constraint foo impossible to satisfy. Now the example gives y = 7 Solution determined by presolve. y = 0.5 Notice that y has been projected onto the interval [0,.5] implied by constraint foo with x == 3. 19980714 Upon encountering end-of-file after at least one variable declaration but no objective or constraint declarations, print "No variables used" rather than "No variables declared". Fix a bug (possibly causing a fault or surprising output) with solexpand's handling of constraints introduced by linearization of piecewise-linear terms. Fix bug introduced 19980615 causing a fault in referencing _VARS. Fix a bug (sometimes invisible, e.g., to our regression testing) in the changes of 19980309 in the handling of nonconvex piecewise-linear terms. Rather than faulting, issue an error message on attempts to assign a symbolic value to a suffix for which no $suffix_table exists. Example: var x; suffix zzz symbolic; let x.zzz := 'abc'; 19980717 Fix fault in expand commands involving a failed consistency check. Example: param p := 3; param q := 2 >= p; var x >= q; minimize zot: exp(x); expand; # faulted after the error message about q not >= p Do not terminate compound commands when a command, such as solve, incurs infeasibility (or other) warnings from presolve. Have the solve command set solve_result_num to 299 (rather than 200) when presolve determines that the problem is inconsistent. 19980716 Fix bug (possible fault or worse) in suffix declarations (including those by the solve and solution commands) issued after a delete command. Minor cosmetic tweak: when echoing new suffix declarations issued by a solve or solution command, only put an extra newline in front of the first echoed suffix declaration. 19980722 For symbolic suffixes, binary or integer applies to the _num value. Fix bugs in "display _scon.suffix, _svar.suffix;" for some builtin suffixes (e.g., _scon.dual when some constraints have been eliminated by presolve). Fix bugs in handling multiple objectives: objectives have always been permuted so nonlinear objectives come first, but the solve command's "Objective =" message did not reflect this permutation, nor did the recording of objective.result and objective.message, nor did some _sobj suffixes. Fix bug (possibly leading to a fault) in assignments by the let command of _obj[i].declared_suffix and _sobj[i].declared_suffix. New builtin suffixes .no and .sno: if foo[i] is _xxx[j], then foo[i].no = j (for xxx in {"con", "var", "obj"}); similarly, if foo[i] is _sxxx[j] then foo[i].sno = j, and foo[i].sno = 0 if the solver does not see foo[i]. Permit "display _obj.user_defined_suffix;" and "display _sobj.user_defined_suffix;". 19980730 Fix bug (fault) under "option times 1" in handling input with line numbers from the C preprocessor when a file ends in an incomplete command. Fix bug in "solution" commands that print options (reflecting that the solution was computed under a different set of options relevant to presolve): presolve was not run again. Work-around: reissue the solution command. Fix possible fault in solexpand (of objectives, when presolve has eliminated one or more constraints) introduced in 19980722. Fix bug (possible fault) sometimes seen in problems with defined variables and piecewise-linear terms. The bug only bit if linearizing the piecewise-linear terms increased the number of constraints to or beyond the next power of 2. Fix some bugs with _scon[...].builtin_suffix for constraints added to linearize nonconvex piecewise-linear terms. Record settings of $linelim and $substout in .nl files, and adjust these options, if necessary, when reading solutions. 19980805 Fix bug (fault) in computing reduced costs for problems with defined variables and no objectives. Correct glitches in the (temporary) treatment of $substout values 2 and 3. Permit presolve to fix defined variables even when this leads to anonymous equality constraints. For now, at least, turning on the "4" bit of $substout inhibits this new behavior. Under option substout 1, this behavior turns out to have been occurring already with potential defining constraints. For example, var x; var y = x; s.t. foo: y = 3; now gets "Solution determined by presolve." whether $substout is 0 (default) or 1. Previously under default conditions, the problem was transmitted to the solver. Now this only happens if $substout is 4. Fix bug in "redeclare var ..." and "redeclare s.t. ..." that could lead to an infinite loop (or worse) upon a subsequent attempt to delete the var or constraint in question. 19980826 Fix bug with delete and redeclare of an entity A that depends on an entity B only referenced by A, after deletions of other entities that depended on B: after subsequent declarations of two or more entities that depend on B, chaos could ensue. Fix bug introduced in 19980722 that caused "ordered by" or "circular by" or "within" phrases in subscripted set declarations to be ignored by "let" commands assigning to subscripts that hitherto had no value; the bug similarly caused restrictions declared on subscripted parameters to be ignored. Example: set S ordered := 2 .. 11; set T {1..2} ordered by S; param p{i in 1..2} >= i; data; set T[1] := 8 3 5; param p[1] 7; let T[2] := {2, 9, 6, 7}; let p[2] := -1; display T; # T[2] not ordered correctly display p; # no complaint about p[2] Fix fault in the sequence sequence "solve; let ...; display _con;" when "let ...;" changed some set or param, so some part of the problem must be reinstantiated, and when presolve had removed one or more constraints from the original problem. (The fault arose whenever something required dual values for the removed constraints to be computed.) 19980828 Fix bug in handling recursive parameters: if, due to conditional expressions, a recursively defined param was not actually used and an update would require recomputing the param (if it had been used), a fault ensued. Example: set S ordered; set T; param p{i in S} := if i == first(S) then 1 else p[prev(i)] + 1; param q{T}; param r{i in S} := if i in T then q[i] else p[i]; var x{S}; minimize zot: sum{i in S} (x[i] - r[i])^2; data; set S := a b c; param :T: q := a 2.4 b 4.2 c 8.9; solve; display x; let S := S union {'d'}; solve; display x; # fault 19980905 Adjust "display x;", where x is an indexed variable, so it does not run presolve. Fix a bug with "drop constr[...];" (dropping some but not all subscripts of an indexed constraint) when defined variables were declared after the constraint. The defined variable declarations were often fairly harmlessly treated as ordinary constraints, but sometimes an incorrect problem resulted from incorrect variables being substituted out. Fix a bug with the sequence solve; var new_var{...}; let{...} new_var[...] := ...; Sometimes storage got corrupted (depending on the original number of variables and the size of new_var's indexing set). Fix a fault in problem foo: something, ...; fix something; # or unfix or drop or restore delete foo; # the current problem problem foo: ...; # fault 19980912 Correction to 'Adjust "display x;"' in 19980904 above; if x was a defined variable, sometimes it was not recomputed. Sat Sep 12 08:40:47 EDT 1998 Fix glitches in "let" commands that result in certain error messages, such as about sets having different arities: the next command sometimes saw the wrong list of dependencies or (with input from stdin) was sometimes erroneously suppressed. Fix bug deleting the current problem under certain circumstances, as in problem zap; var x; minimize zot: (x-1)^2; solve; delete zap; # fault Fix possible fault with multiple solves when a nonlinear defined variable is fixed by presolve. 19980928 Fix bug with "display _scvar;", which showed incorrect values. Fix bug (fault) with solexpand complementarity_constraint, as in var x; var y; test: x >= 0 complements x + y <= 3; solexpand test; Fix bug (fault) with "expand _scon;" when presolve has resolved all complementarity conditions. 19981005 Fix possible fault with "let" assignments to generic synonyms, such as _var and _con. If the fault didn't occur, the bug was harmless. 19981009 Fix fault with sets having a random default expression. Example: set A default 1 .. 10*Uniform01(); display A; # fault Permit ":= {}" in set declarations. Permit set expressions of the forms if test then {} else set_expr and if test then set_expr else {} with the set_expr supplying the dimen (arity) of the empty set {}. 19981014 Permit "display p.result;" for indexed collections of problems p. For infeasibility detected by presolve, have the solution, solve, and write commands update the 'result' suffix of the current problem. 19981019 Fix glitch in distinguishing variable.status and variable.astatus values "fix" and "unused". Fix rarely seen bug that resulted in the error message "invalid refct 0 in opgen". 19981022 Fix possible fault in computing var.lrc, var.urc, constr.ldual, and constr.udual when the problem has no objective (i.e., is just the problem of finding a feasible point). 19981105 Fix bug in handling imported functions: with some sequences of commands, incorrect .nl files were generated, leading solvers to fault. Example: function foo; var x; s.t. Foo: foo(x) >= 0; let Foo := 1; # initial dual variable solve; # fault In this example, an easy work-around is to provide the initial dual variable value in the constraint declaration: s.t. Foo := 1: foo(x) >= 0; or in a data section: data; var Foo := 1; Fix glitch in handling certain problems in which presolve detected infeasibility: a second solve command sometimes sent a feasible problem to the solver. Example: set I := 1..2; var x{I} >= 0; minimize zot: sum{i in I} i*x[i]; s.t. fixit: x[1] = 0; s.t. bletch: x[1] = 2; 19981109 Fix bug in parsing some "setof" expressions and literal sets involving tuples of expressions: the initial components of the tuples were lost when the last was a string expression. Examples: display setof{i in 1..3} (i, sub('123',i,'x')); display {i in 1..3}: {(i, sub('123',i,'x'))}; Fix glitch with single stepping: a single-step command involving a bad subscript could terminate execution. New builtin params recording execution times (where possible): _solve_time = _solve_user_time + _solve_system_time record CPU seconds for the most recent solve command, and _solve_elapsed_time records wall-clock seconds for the most recent solve; prepending _total gives sums of the corresponding entities for all solve commands; changing "solve" to "shell" gives analogous values for shell commands. And _ampl_time = _ampl_user_time + _ampl_system_time and _ampl_elapsed_time give analogous times for AMPL itself, excluding times for shell and solve commands. On all systems, the elapsed time params are computed by the time() function, which has a granularity of whole seconds. The new CPU time params report true CPU seconds on Unix and Windows NT systems (with Win32 binaries, such as ftp://netlib.bell-labs.com/netlib/ampl/student/mswin/samplnt.exe.gz). On Windows 9x or 3.x, and on MS-DOS systems, CPU times are not available; the system times are set to 0, and user times are elapsed (wall-clock) seconds since the start of execution -- with a finer granularity than the time() function. 19981120 Fix bug (fault or worse) in handling suffixes declared with restrictions on their values, such as inequalities or "integer": after an assignment to the suffix, references within an iterated context might encounter the bug. Example: set I := 1..3; var x{i in I} integer; suffix foo >= 0 <= 10; let{i in I} x[i].foo := i; let{i in I} x[i].foo := x[2].foo + 1; # fault (If "2" were changed to "i", the fault would go away. Inserting "display x.foo;" before the second "let" would also bypass the bug.) Fix possible glitch in scaling of CPU times on some systems (that do not have getrusage()). Fix bug in handling models with nonlinear defined variables that appear in piecewise-linear terms. Under unpredictable conditions, an invalid .nl file sometimes resulted. Unless option linelim 1 (or command-line option -L) is specified, all defined variables appear to be nonlinear; specifying option linelim 1 may thus avoid the bug. Fix fault (or worse) in references to _var.defeqn or _var.dual after presolve has found the problem infeasible. Fix glitch in _con.defvar: under unusual circumstances, some components were erroneously reported as nonzero, e.g., when option substout 1 was in effect and there were two or more candidate defining constraints for a variable. 19981214 Adjust error message for a symbolic data-section value for a numeric parameter. For example, the old message type['HOST'] cannot be #ENDPOINT is now type['HOST'] must be numeric, not 'ENDPOINT' Adjust "objective;" command so it does not try to instantiate the current problem unless there is a partially dropped array of objectives (in which case ampl must worry about which subscripts are valid). "Invisible" change: put suffix information ahead of nonlinear information in .nl file, so .nl reading can be influenced by suffixes. Fix seldom-seen bug in invocations of the form ampl -ogfoo boo2 in which something recursive or computed from a default expression involves something indexed over a set whose definition involves subscripts: part of the subscript computation may have been latched prematurely, causing subsequent failure. Here is "boo2" of the offending example: param n{1..2}; set IR := {'i','r'}; param f{i in 1..2, 0..n[i], IR}; var a{IR}; var phi{i in 1..2, j in 0..n[i], ir in IR} = f[i,n[i]-j,ir] + ( if j == 0 then 0 else if ir == 'i' then a['i']*phi[i,j-1,'r'] + a['r']*phi[i,j-1,'i'] else a['r']*phi[i,j-1,'r'] - a['i']*phi[i,j-1,'i']); minimize N: (phi[1,n[1],'i']^2 + phi[1,n[1],'r']^2) / (a['r']^2 + 1); data; param n := 1 2 2 2; param f default 0 := [1,*,r] 2 1 1 -3.001 0 2.01 [2,*,r] 2 1 1 -4.001 0 3.0002 ; var a := r 1.001 i 0; 19981218 Fix bug (possible fault, or worse) in reading solutions of problems having variables or constraints declared with {if false_condition}. Forbid assignments to such variables and constraints. Include in the $eexit processing error messages for suffixes not assigned values consistent with restrictions in the suffix declarations. Fix glitch in passing initial guesses: before the first solve, "let" assignments to _var were sometimes ignored (in favor of the initial guess in the variable's declaration or in a data section). 19981230 Fix fault in handling constraints or objectives with {if true} indexing and piecewise-linear terms. 19990112 Fix bugs with some "redeclare ..." declarations that changed the type of the declared entity. Example: param p := 3; redeclare var p; minimize zot: (p-1)^2; # error message Fix bug with "shell (string_expression);": derived quantities in the string_expression were not updated. Fix glitch in error message for "update data zork;" when zork is not defined: instead of "s is not defined", say "zork is not defined". Fix possible glitches with "show su; show e;" after a reset. Ignore gratuitous reset commands. Modified linearization of piecewise-linear terms. To get the linearization that had been the default since 19980309, specify option pl_linearize 3; 19990120 Fix glitch (possible fault, or worse) in the handling of "hard" (nonconvex) piecewise-linear terms in version 19990112. New addition to reserved-word list: table. 19990129 Fix bug in command-line invocations of "ampl -o..." when the problem involves imported functions: an invalid .nl file resulted. 19990202 Fix (reference-counting) bug in references to any of the sets Reals, Integers, ASCII, EBCDIC, and Display. The bug manifested itself in the second command to reference a particular one of these sets. Example: print if 'A' in Reals then 'yes' else 'no'; # works OK print if 'A' in Reals then 'yes' else 'no'; # error or fault Fix bug with "redeclare": declaring a new entity, such as a new variable, then redeclaring another entity to involve the new one could lead to a subsequent fault in using the redeclared entity if it had indirect dependencies. Example: model diet.mod; data diet.dat; var x; redeclare minimize total_cost: sin(x) + sum{j in FOOD} cost[j]*Buy[j]; write 0; # fault 19990209 Fix possible bug with reading (initial guess) data for variables or constraints in a data section after "update data;" and after the variables or constraints have been instantiated. The bug would bite only under unusual conditions and would have unpredictable consequences. ftp://netlib.bell-labs.com/netlib/ampl/student/mswin/samplnt.exe.gz: fix an infinite loop (due to a compiler bug) in evaluating random functions, such as Uniform01(). 19990225 Fix a bug in an unadvertised feature: permitting subscripts after some dot notations rather than within them. If x is a subscripted variable, then references to, say, x.lb[i] (or any other built-in suffix other than sstatus) worked, but x.sstatus[i] and x.any_declared_suffix[i] could fault. Fix bug (possible erroneous "not generated" message) in the let command's handling of assignments to problem.declared_suffix. Fix bug with "display thing.suf_num" -- "_num" was omitted from the column heading when "suf" was a declared suffix. 19990304 Obscure bug fixes: when reading files (e.g., stdin under Unix with manual control-Ds giving partial lines, or in general with very long input lines), when it appears that a token may be split across two read() invocations, do not ignore white space (which terminates the token) at the start of the second read(); and when reading stdin, give the secondary rather than the primary prompt before the second read(). ftp://netlib.bell-labs.com/netlib/ampl/student/mswin/samplnt.exz: Under W9x, treat '/' as '\' in the names of programs invoked by the shell and solve commands. (This already worked under NT, but Microsoft does not tolerate much consistency among its operating systems.) 19990305 Fix glitch in yesterday's change: a comment at the end of a file file gave rise to the wrong prompt just after an "include" of the file. 19990310 Fix glitch in handling (final) files that do not end with newline. 19990311 Fix botch in yesterday's change affecting two or more command-line files. 19990323 Fix bug (possible subsequent fault, or worse) with redeclaring imported functions. Fix another recently added bug reading files without a final newline. Turn {set_expr} into set_expr. Fix glitch in prompt after "reset;" when there have been no declarations but there was a syntax error in a compound command. 19990326 Fix a memory fault that arose under obscure conditions; absent the fault, the bug was harmless. In part, the bug involved "reset data p;" after assigning some values to a parameter p but never using them. For the record, here is an example where the bug bit: set A; set B := {i in A: i != 2}; param p{A} >= 0; let A := 1..2; let{i in B} p[i] := i; let A := 3..4; display B; reset data p; let{i in B} p[i] := i+2; Fix a bug with reading ASCII format .sol files from files with carriage-return characters (e.g., text files on Microsoft systems). 19990421 Fix possible fault in command-line invocations ampl -ogfoo foo.mod foo.dat ... when presolve determines all the variables. Fix a bug that led to error messages of the form "presolve has k = 0, P.nfc = 1" (under complicated and rare conditions). Try to avoid leaving temporary files behind in the face of various signals, such as SIGINT and SIGHUP. Have printf "%q",''; print '' rather than nothing. New builtin symbolic param _cd is set to the current directory initially and each time the cd command is executed. New variant -bs (and -brs) of command-line option -b for GUIs that want to run solvers themselves. Ask dmg for details. Have "option solver_msg 0;" suppress messages that appear when presolve determines the solution (as well as the solution_message returned by a solver). 19990426 Fix bug with "option linelim 1;" in models with partially linear defined variables that (e.g.) appear both linearly in another defined variable that is only used linearly and nonlinearly in a constraint or objective: displaying the partially linear defined variables led to an invalid .nl file. Example: var x{i in 1..4} := i; var s = sum{i in 1..4} i*x[i]; var y{i in 1..2} = x[i]^2 + s; var z{i in 1..2} = y[i] - 5; minimize zip: sum{i in 1..2} (3*z[i] + y[i]^2); option linelim 1; write gzap1; # correct display y; display z; write gzap2; # was bogus 19990513 If SIGHUP is being ignored (e.g., by nohup), continue to ignore it. Fix glitch with option show_stats 1: in problems with nonlinear objectives, the "nnn nonlinear objective(s)" line erroneously mentioned "linear" nonzeros. 19990607 Table and out-arg extensions, to be described later. Fix glitch in "show" command: the example param N; set S; var x{1..N,S}; var y{1..N}; minimize zot: sum{i in 1..2} y[i]*sum{j in S} x[i,j]; data; param N := 2; set S := a b; write 0; show x; printed "var x{y.index, S};" rather than "var x{1 .. N, S};". Fix a bug that could only bite on Intel processors: under unusual conditions, nonlinear defined variables sometimes incorrectly caused generation of NaN values in or after a "solve" or "solution" command. 19990804 Fix fault in the sequence option log_file 'foo'; var x; reset; option log_file 'foo'; var x; reset; (i.e., in issuing a redundant "option log_file" command after a nontrivial reset). Fix bug in handling /* ... */ comments longer than about 4300 characters: they were inadvertently treated as end of file. Change "amplodbc.dll" to "ampltabl.dll". Fix bug in "delete suffix foo;" and "delete table foo;" that could cause memory corruption. When an input file ends with a #comment not terminated by a newline, assume a newline after the comment. Permit the close command to mention a comma-separated list of files to close. Change to filename syntax: disallow commas in unquoted filenames. This permits one to say things like load lib1, lib2; and close file1.stuff, file2.xyz; without treating the "," as part of the file name. Change to $ampl_include: previously it was a white-space separated list of directories in which to look for files mentioned by "include", "model", and "data"; now spaces are allowed within directory names (but not before or after them), and directory names in $ampl_include should be separated by newlines (or tabs, but tabs invite confusion; for example, X-Windows turns tabs into spaces). Change $AMPLFUNC to behave similarly to $ampl_include: directory and file names may contain internal blanks, and $AMPLFUNC now specifies a sequence of zero or more file or directory names from which to import functions. Each name is first treated as the name of a shared library (or DLL -- suffixes, such as .dll, must be explicit); if a library by that name cannot be opened, the name is treated as a directory name, and "/amplfunc.dll" (or "\amplfunc.dll" under MS Windows) is appended to obtain the name of a library to load. The solve command now adjusts $ampl_funclibs to be a newline- separated list of full pathnames of libraries from which functions in the current problem instance were imported. New commands load [libname [, libname ...] ]; unload [libname [, libname ...] ]; reload [libname [, libname ...] ]; load, unload, or reload shared libraries (from which functions and table handlers are imported). When at least one libname is mentioned in the load and unload commands, $AMPLFUNC is modified to reflect the full pathnames of the currently loaded libraries. The reload command first unloads its arguments, then loads them. This can change the order of loaded libraries and affect the visibility of imported functions: the first name wins. With no libname arguments, "load;" loads all the libraries currently in $AMPLFUNC; "unload;" unloads all currently loaded libraries, and "reload;" reloads them (which might be useful if some have been recompiled). New and adjusted system sets: _LIB = currently loaded libraries; _AVAILFUNCS = currently available imported functions (which still must be declared before use) _AVAILFUNC2 = (available function, library) pairs (permits seeing which library is currently supplying an available imported function) _SFUNCS = imported functions that "solve;" currently uses. New command remove [filename [, filename ...]]; closes and removes the files mentioned. "remove;" does nothing. Exception: "remove ($log_file);" just truncates (removes and then reopens) $log_file, much as "close ($log_file);" does. Extensions of funcadd.h for (still to be described) table handlers; the updated funcadd.h is now in /netlib/ampl/solvers. 19990806 Fix bug in handling ampltabl.dll (intended for use with the table extensions, still to be described). Fix glitch with "write 0; solve;" -- "solve;" now writes a .nl file. Note that "write 0;" sets $ampl_funclibs. If a reference to _SFUNCS requires it to be recomputed, $ampl_funclibs is now also updated. Fix bug with reloading libraries that provide at least one already- imported and declared function: a fault was then possible with "display _AVAILFUNC2", "display _AVAILFUNCS", and "show functions". Fix bug that sometimes led to a surprising error message involving "because its declaration had no indexing set_name" during parsing of table declarations. Example: table T : [i=ABC,j=DEF] a[j,i]; (Note that with the builtin table handlers and sample ODBC handler, this declaration could be written more simply as table T : [DEF,ABC] a; since these handlers permit the names in table declarations to be in a different order than those in the external table.) Fix bug with "update data" or "read table": adding values to, e.g., a param declared without any consistency checks could lead to a fault. Example: set B; param q{B}; data; set B := a b c d; param q := a 10 b 20; display q; update data q; data; param q := c 30 d 40; display q; #faulted Fix bug in discarding bad subscripts under some conditions. Example: set A; param p{A}; param q{A}; data; param :A: p q := a 1 10 b 2 20; display p,q; update data q; data; param q := c 30 d 40; display {i in A} (p[i], q[i], p[i] + q[i]); #faulted New system param _version gives the AMPL version number. 19990811 Fix longstanding bug in handling problems involving subsets of indexed constraints, such as the stoch2 problem in http://www.ampl.com/ampl/NEW/LOOP2: sequences of the form solve prob1; # involving only some subscripts of a constraint let ...;# evaluate something not previously instantiated # that also involves dual values of the partially # dropped constraint solve prob2; # involving different subscripts of the constraint resulted in wrong dual values subsequently being used in expressions using dual values of the indexed constraint in question. The bug apparently crept in sometime between 19960823 and 19961129. 19990812 --> 19990813 Names "sampl" and "samplnt" in /netlib/ampl/student/* changed to "ampl"; /netlib/ampl/student/sun4 changed to /netlib/ampl/student/solaris. The old names will continue to work for a few more weeks; they're linked to the new ones. 19990818 Fix a fault in handling incorrect -b command-line options. Fix bug, introduced in 19990804, in for{i in S} {...}: if i (the dummy variable of the "for" loop) was all that changed in {...}, then various things didn't get recomputed, leading to surprising results. Example: set S{1..3}; for{i in 1..3} let S[i] := i..i+3; display S; gave set S[1] := 1 2 3 4; set S[2] := 1 2 3 4; set S[3] := 1 2 3 4; Since the "let" command can be indexed, there was no need for a "for" loop in this example, which would better have been written set S{1..3}; let {i in 1..3} S[i] := i..i+3; display S; which gives the expected output set S[1] := 1 2 3 4; set S[2] := 2 3 4 5; set S[3] := 3 4 5 6; Making something else change in the for loop also resulted in correct behavior, which explains why we didn't see this bug earlier. For instance, set S{1..3}; param j default 0; for{i in 1..3} { let S[i] := i..i+3; let j := j + i; } display j, S; worked correctly. MSDOS and Win32 "ampl" binaries: add expansion of * and ? in command- line arguments, and replace command-line arguments of the form @filename with the contents of file filename. Win32 "ampl" binaries: recognize quoted white space in command-line arguments, no matter who provides the command line (e.g., another program calling CreateProcess). White space may be quoted with either single or doubled quotes (' or "). Within a quoted string, the quote character may appear if it is doubled (as in AMPL strings). 19990821 Fix some bugs with the interaction of "load" and "reset". 19990830 Fix glitch with "objective;", which sometimes responded with a sequence of "drop" and "restore" commands instead of a single, equivalent "objective" command. Example: var x; minimize foo: sin(x); minimize goo: cos(x); objective foo; objective; 19990831 Fix bug in "redeclare" of entities on which other entities depend. Declaring a new entity that depended on the redeclared one caused memory corruption. Example: model diet.mod; redeclare set FOOD ordered; data diet.dat; param T{j in FOOD} := 199900000000 + ord(j)*1010101.1; solve; # fault 19990908 Some changes to "tables", an extension we hope to describe soon. 19990913 Fix glitch in parsing name.suffix[subscripts]: an internal array was corrupted in a way that could cause subsequent surprising behavior. For example, a reference to constraint.dual[subscripts] in a loop or {command; command...;} block could cause the constraint to be dropped in the first solve. Example: var x{i in 1..3} >= 0; s.t. c{i in 1..2}: x[i] + x[i+1] = 1; minimize foo: sum{i in 1..3} i*x[i]; { # start block solve; display x,c; print{i in 1..2}: 'c[',i,'].dual = ', c[i].dual; print{i in 1..2}: 'c.dual[',i,'] = ', c.dual[i]; } # Printed x = c = 0; c.dua[i] in the last printf caused # the trouble. With the last printf commented out, correctly # printed x = 0,1,0 and c = -1,3. 19990916 "Invisible" portability tweaks to binary <--> decimal conversion routines. 19990930 Fix data-section glitch: tables of the form param :setname: ... := ... ; that supply both parameter values and an indexing set didn't record that setname had been assigned a value. This permitted set A; param p{A}; param q{A}; data; param :A: p := a 1 b 1.5 ; param :A: q := x 3.3 y 4.2; display A; to get by without an error message. (A subsequent "display p;" did produce an error message about invalid subscripts for p.) Fix bug in turning {..., (a,b) in {A,B}, ...} into {..., a in A, b in B, ...}: under apparently rare conditions, unpredictable behavior was possible; e.g., slices over A or B could go awry. 19991004 Fix some glitches with table declarations and "write table". 19991006 Fix a nasty bug that crept in sometime in the latter half of 1995: use of a slice in a command could foul up subsequent uses of the same slice in subsequent independent commands (not part of any compound command containing the first use). Example: in set A dimen 2; data; set A := a b a c b c x z; display {('a',i) in A}; display{i in 1..5} i^2; display {('a',i) in A}; the final display incorrectly listed 1 2 3 4 5 rather than b c. 19991027 Fix bug with "read .. < -;" introduced around 19960401 that could cause a fault or other unpredictable behavior. Fix bug in handling variables subscripted over empty sets after a solve, subsequent changes to entities in the variables' declarations, and another solve or something else that caused the variables to be regenerated. Example: set S default {}; param p default 0; var x{S} >= p; var y; minimize zot: (y-1)^2 + p*sum{i in S}x[i]; solve; let p := 2; solve; # unpredictable behavior was now possible Fix a bug (fault) in handling duplicate appearances of the same slice, as in param p {w in W, j in 1..n} := if card {(nw,w) in P[j]} = 1 then sum {(nw,w) in P[j]} nw; Fix bug in computing the value of a piecewise-linear function of a defined variables within an AMPL session (e.g., in a display command, but not in expressions sent to solvers): if nothing else had caused the defined variable to be evaluated, the piecewise-linear function could be miscomputed. Example: var x := 2; var y = 3*x + 1; var z = <<3; -1,1>> y; print z; # printed 0 rather than 1 print <<3; -1,1>> y; 19991130 Fix bug with redeclare: redeclared entities were sometimes not regenerated. Example: set I := 1..299; var x{I} binary; maximize zot: sum{i in I} x[i]; s.t. knapsack: sum{i in I} Uniform01()*x[i] <= 43; solve; # objective 159.0316356 redeclare set I := 1..300; solve; # gave same objective before bug fix # after fix, objective 162.1322059 Fix bug with if expressions in nonlinear expressions: in some (complicated) situations in which the test involves only "outer" dummy variables and the "then" or "else" parts involve inner dummies, a fault was possible. Example: set A; set B; set C; var x{A, B}; var z{B} >= 2; var y{i in A, j in B} = z[j] * (if i in C then 0 else (x[i,j] - 2)^2); minimize zot: sum{i in A, j in B} y[i,j]; data; set A := a b c; set B := x y; set C := b; Win32 binaries only: adjust the cd command so sequences like cd c:/some/place/other/than/the/starting/directory; cd d:; print 'something' >'c:zot'; shell; will show the adjusted current directories to spawned processes (in "solve" and "shell" commands) and will interpret 'c:zot' correctly (as c:/some/place/other/than/the/starting/directory/zot). Note that under DOS, W9x, and NT, you can use / in place of \ in file names. [For reasons known only to Microsoft, under W9x, a spawned command.com gets its current directory information in some other way than do Win32 programs, so "shell" commands under W9x only reflect cd commands for the current drive. This problem goes away if you set option COMSPEC to a Win32 command processor.] 19991207 Fix a bug in presolve's handling of complementarity conditions that led to error messages of the form "presolve has k = 5, P.nfc = 4". Workaround: option presolve 1; example: var x1; var x2; s.t. c1: x1 + x2 >= 2; s.t. c2: x1 - x2 <= 0; s.t. c3: 0.1*x1 + x2 <= 1.1; s.t. c4: x1 >= 0; var y{1..2}; s.t. foo: y[1] + y[2] >= 3 complements x1 <= 1.02; s.t. goo: sum{i in 1..2} (y[i] - i)^2 >= 2; 20000128 New command-line option -vi[nnn] to specify whether stdin should be treated as interactive. The default is to choose based on isatty(): if stdin and stderr both appear to be a terminal, i.e., isatty(0) and isatty(2) both return 1, assume stdin is interactive. Invoke ampl -v? for a summary of all -v options; the description of -vi is -vi[nnn] {interactive mode? (Must be first.) nnn = 0 ==> no, nnn = 1 ==> yes but no prompts, nnn = 2 ==> yes with prompts, nnn = 3 ==> 0 or 2, based on isatty()} In non-interactive mode, syntax errors now inhibit execution of commands and terminate reading of the current file; previously, commands were sometimes executed. (Block input mode, used by AMPL Plus and sample GUIs in http://www.ampl.com/ampl/GUI/expermt.html, remains interactive.) New default presolve behavior: simplifications that appear to convert nonlinear expressions to linear expressions really do so. Change to option linelim: "option linelim 0" and "option linelim 1" work as before, while "option linelim 2" and "option linelim 3" suppress the new conversion of "nonlinear" linear expressions resulting from presolve simplification to true linear expressions, and otherwise behave like "option linelim 0" and "option linelim 1", respectively. Fix theoretically possible glitch in handling "break loopname;" and "continue loopname;". Whether the glitch could cause trouble is system-dependent; most likely, it would never cause trouble. In the "context: ..." portion of error messages, omit initial white space. Fix bugs with "load" and "reload": if a function in the loaded library had been declared inconsistently with the version in the library, the resulting error message omitted the function's name and could fault; if the library's funcadd called at_reset or at_exit and all its addfunc were rejected, a fault was possible when the functions passed to at_reset or at_exit were finally called (since the library may have been unloaded). Now these functions are called immediately if everything else in the library was rejected. Ignore errno settings by imported functions, which should assign al->Errmsg if something goes wrong. In functions imported by a load command, calls on fprintf should no longer cause an "Error executing..." message. Fix obscure presolve bug in handling constraints or objectives involving a variable that presolve fixes and a nonlinear mod expression whose left operand starts with unitary minus. Example (which looped): var x; var y; minimize zot: exp(-x mod .3) + y; s.t. nail_y: 1 = y; Fix glitch in handling multiplications by (-1) in some situations where it would be cleaner to write a simpler expression. Example: var x; minimize zot: -x*-1; solve; gave "invalid refct 0 in opgen". Fix possible fault on machines with 64-bit pointers with "expand;" and "solexpand;" of models with piecewise-linear terms. 20000204 Fix some glitches with tables (an extension to be described soon). 20000207 Fix botch (possible extra text in $AMPLFUNC) in "unload" and "reload". Prevent some builtin params (such as _ampl_user_time) and sets (such as _LIBS) from being assigned. Fix bug in the interaction of "close;" and $log_file (that could cause "writing" to a closed file). 20000208 Fix another glitch with "unload ...;". 20000209 Fix a glitch with tables. 20000216 Fix bug in linearizing nonconvex piecewise-linear terms: if presolve deduced bounds that permitted eliminating sufficiently many constraints with "hard" (nonconvex or nonconcave, as appropriate) piecewise-linear terms, a fault or unpredictable behavior could result. Increase the resolution of _solve_elapsed_time, _shell_elapsed_time, and "option randseed 0;". The latter now changes more quickly. Warn about "fix x := 3;" when x is a subscripted variable. New builtin LOCAL suffix .relax: for integer or binary variables, .relax > 0 indicates that the integrality of the variable should be ignored. Turn <<0;0,0>>x into 0. 20000217 Omit an explicit "+ 0" that was in the .nl file for var x; var y; minimize zot: sin(x+y+sum{i in 1..0} (x*y)^i); Fix glitch in yesterday's .relax implementation: reading .relax values from a table or OUT args in a function call did not necessarily affect the next "solve". Arrange for integer_variable.relax values >= 2 to cause solvers to see the variable as continuous, but for presolve to treat it as an integer variable. Also, have nonzero $relax_integrality to take precedence over .relax suffix values, so option relax_integrality 1; # ignore integrality everywhere and option relax_integrality 2; # use integrality in presolve, # but tell solvers all variables # are continuous operate as heretofore, regardless of .relax settings. For "let" assignments to variable.declared_suffix, do not require the "collect" phase. 20000228 Under MS Windows, allow for spaces in $TMPDIR. In situations where bounds on a defined variable are equivalent to bounds on a problem variable that the solver sees, reflect bounds on the defined variable to the problem variable and remove the constraint that implied the bounds on the defined variable. Example: var x; var y = 3*x + 2; s.t. ybound: 5 <= y <= 11; # reflect into bounds on x minimize zot: y^2; Deduce bounds on defined variables in more cases. Temporary(?) treatment of $substout: the "4" bit of $substout, i.e., arranging that ($substout mod 8) >= 4, suppresses today's reflection of bounds and stronger deductions of defined-variable bounds. Distinguish singular and plural in more "option show_stats 1" output. Fix glitch with variable.defeqn and constraint.defvar that gave "0" for linear defined variables. Fix bug in handling synonyms: under complicated conditions, a fault was possible. Fix glitch with "solexpand;": if some members of an indexed family of constraints were dropped, others in the family were sometimes shown as "dropped" and expanded in the modeler's view rather than the solver's view. Fix glitch in conversion of "nonlinear" linear expressions introduced 20000128: constant terms were sometimes botched. 20000307 Fix minor glitch with handling a syntax error that terminates reading a data section: the error message could show too much text. Fix possible fault with "display foo;" when foo is indexed by a set involving a set-valued "if" expression. Fix glitch with $presolve_warnings: prior warnings had counted against $presolve_warnings; now only warnings during the current invocation of presolve count. Arrange for "option presolve_warnings -1;" to suppress both the presolve message about how many presolve messages were suppressed and the message about which modified options might help. Plug a memory leak (somewhat unlikely with default settings) that occurred when exceeding |$eexit| warnings during presolve caused presolve to be aborted. 20000309 Fix minor printf glitch: printf "%.0f\n", .1; printed "0." rather than "0" -- under "%.0f", numbers less than 1 in absolute value that do not round to 1 or -1 should print as "0" with no trailing decimal point. Fix possible infinite loop in cleaning up after "Abandoning compound command to accept declaration of ...". If input was directly or indirectly from stdin (via "include", "model", etc.), return to stdin for further input; otherwise quit. 20000316 Fix bug introduced 20000216: suffix declarations appearing after any variable, constraint, objective, or problem declarations caused internal inconsistency, sometimes resulting in a fault when needed entities were not instantiated. Fix longstanding bug: some cached values were not updated after presolve changed relevant variables values. (They were updated after "let", "solve" and "solution" commands, which explains why we didn't notice this sooner.) For example, print _obj[1]; write 0; print _obj[1]; showed the same value twice, even if presolve changed some relevant variable values. Fix a glitch in handling complementarity conditions: if presolve detects infeasibility, adjust things to avoid the mysterious error message "presolve has k = ...". Fix longstanding bug with iterated "let" assignments to a subscripted param with a default value and a right-hand side that involves the left- hand side: if the default value had to be evaluated when the number of already assigned subscript values was a power of 2, chaos could ensue. Increase the resolution of _ampl_elapsed_time, and have "reset;" reset _ampl_elapsed_time, even when there were no declarations to discard. (Otherwise, "reset;" is a no-op in this case.) 20000317 Fix glitch introduced in the first "correction" of 20000307: param p; data; param p := 3;display p; and param p; data param p := 3; model; display p; (transitions from data mode to model/command mode not begun at the start of a line) were mishandled. 20000321 Fix bug with "option substout 1" introduced 20000128: a fault was possible, e.g., var x >= 0; var y; s.t. ybounds: 5 <= y <= 11; s.t. ydefn: y = 3*x + 2; option substout 1; solve; # fault 20000327 Tweak to hash-table logic to speed up reading data on some large examples (with many symbols). Correct glitch with complementarity constraints that caused some linear constraints to appear nonlinear when involved. Fix longstanding bug with option linelim 1: if a constraint or objective made nonlinear use of a nonlinear defined variable that depended linearly on certain other variables, and if these other variables did not otherwise appear in the constraint or objective, they were omitted from the list of partial-derivatives that could be nonzero. Example (variant of Rosenbrock test function): under "option linelim 1", variable x[2] was omitted from the lists of possibly nonzero partials for the objective and constraint. var x{1..2}; var f1 = 10*(x[2] - x[1]^2); var f2 = 1 - x[1]; minimize f: f1^2 + f2^2; s.t. zip: f1^2 >= 3; data; var x:= 1 -1.2 2 1; Tweak to "option linelim 1" output: nonlinear defined variables that have no linear terms but are used linearly in a constraint or objective are no longer split into two parts (unless, temporarily, the "4" bit of $linelim is on, i.e., $linelim mod 8 >= 4). Command-line option -L now accepts an optional integer immediately following the "L" (no space) to specify the initial $linelim setting. No integer ==> -L1 (the old behavior). Change the default option linelim from 0 to 1. 20000406 Fix longstanding bug with "option linelim 1" on complementarity problems: adjustments to the constant term in complementarity constraints were missed. 20000412 Fix trouble handling many numeric set members introduced 20000327. Fix longstanding bug that could make set member 3000 misbehave. 20000413 Fix a long-standing bug that could afflict scripts that solve sequences of complementarity problems when memory got corrupted. The "fix" of 20000406 could occasionally lead to corrupted memory; try again... 20000427 Fix bug in providing variable and constraint (dual variable) initial guesses in data sections after the variables or constraints in question had been instantiated: providing an incomplete set of values sometimes led to chaos. Tweak permitting use of AuxInfo (new field ae->AI) in imported functions. This is available when ae->ASLdate >= 20000427. In calls from AMPL, ae->AI is always null, but solvers can arrange for it to have useful values. 20000428 Fix bug in "expand variable;"'s handling complementarity constraints. Example: set I := 1..2; var x{I}; s.t. goo{i in I}: x[i] complements x[i] == 0; expand x; gave Coefficients of x[1]: goo[1].L 1 goo[1].R 1 Coefficients of x[2]: ### None ### Fix bug with handling partially dropped complementarity constraints: incorrect names were sometimes generated, and memory could be corrupted. 20000430 Fix another bug with handling partially dropped complementarity constraints -- another context where memory could be corrupted. 20000502 Fix bug (possible fault or wrong constraint or objective subscripts) with "expand variable"'s handling of partially dropped constraints. Example: model transp.mod; data transp.dat; drop Supply['GARY']; expand Trans['CLEV','FRA']; # wrong Supply subscript Fix another glitch with partially dropped complementarity constraints. 20000505 Fix nasty bug with adjusting nonlinear constraints that become linear due to variables being fixed during presolve: a test to decide which such constraints might imply further bound adjustments was wrong, leading to the possibility of incorrectly eliminating constraints. Work-around: "option linelim 2;" to suppress the adjustments. Fix bug with converting previously nonlinear expressions to linear expressions when the limit $presolve - 1 on extended presolve iterations is reached and the final iteration made more conversions possible. A fault was likely. Temporary fixes: increase option presolve or specify option linelim 2 (to suppress the conversions). Fix some minor memory leaks. Arrange for "option compl_warn 0" to suppress the warning message about nonsquare complementarity systems and for "option compl_warn 2" to cause nonsquare complementarity systems to be infeasible. Fix bug with computing suffix values (such as _con.status) for complementarity constraints. Trouble remains with .sno and .no. 20000512 Fix another bug with converting previously nonlinear expressions to linear expressions: adjustments to constraint bounds due to variables fixed by presolve were made twice, possibly leading to incorrect warnings of infeasibilities or incorrect bounds transmitted to solvers. Fix possible miscomputations of constraint.no and constraint.sno when the constraint is partially dropped. For complementarity constraints, make .Lno, .Lsno, .Lstatus, .Lstatus_num, .Lastatus, .Lastatus_num, .Rno, .Rsno, .Rstatus, .Rstatus_num, .Rastatus, and .Rastatus_num work. Adjust the expand command so empty constraint bodies print as "0". Example: var x{i in 1..2}; s.t. zap: sum{i in 2..1} x[i] == 0; expand zap; now prints s.t. zap: 0 = 0; rather than s.t. zap: = 0; 20000515 Fix bug introduced 20000128 in simplifying nonlinear "min" and "max" expressions involving a constant term. The bug only bit when presolve had eliminated at least one variable. Example: var x{1..2}; minimize zot: (x[1]-7)^2; s.t. blat: x[2] = 3; s.t. silly: max(x[1], x[2], 1) <= 4; # got lost 20000516 Fix longstanding bug in generating nonlinear variables "out of order", under unusual circumstances, as in var w; var x{1..2}; var y{1..10}; minimize zip: (w-1)^2 + sum{i in 1..2} (i+1 - x[i])^2 + sum{i in 1..10} (i+3 - y[i])^2; s.t. zap: sum{i in 1..10} y[i]^2 == 4; let x[1] := 3; let zap := 4; solexpand; in which x is generated before w and y and (it turns out) because of the dual-variable assignment "let zap := ...", y is generated before w. Nonlinear references to y were misnumbered, leading in the above example to an incorrect expansion of "zap" -- and in another example to a fault. Fix glitch with certain dual-variable assignments to hitherto ungenerated constraints, as in var x{1..2}; minimize zip: sum{i in 1..2} (i+1 - x[i])^2; s.t. zap: sum{i in 1..2} x[i] == 4; let x[1] := 3; let zap := 4; The last "let" command elicited an erroneous error message about zap being an ungenerated constraint. 20000528 Fix obscure bug with "option presolve 1;": if a defined variable was fixed by presolve and appeared in an equality constraint with one other variable, then an invalid .nl file resulted. Example: var x{1..2}; var y = x[1] + 1; s.t. zip: y = 1; minimize zap: x[1]^2 + x[2]^2 + y^2; Work around: leave option presolve at its default value (10). 20000602 Fix a memory leak in computing reduced costs. Treat variables in defining declarations (var x = ...) as defined variables even when bounds are reflected on them. 20000605 Fix glitch in, e.g., "display objective.result;" introduced in version 20000512. Example: model dietobj.mod; data dietobj.dat; objective total_cost['A&P']; solve; objective total_number; solve; display total_cost.result, total_number.result; gave "unexpected case -94 in sdotvalue". Fix a longstanding bug that gave an error message of the form "unexpected type xxxx in ecopy()". Example: var x := 3; var y = x*sum {a in 1..1, b in 1..1} if a=1 then x; display x, y; 20000608 Bypass a bug with some compilers (e.g., Microsoft Visual C++) that caused NaN == x to be "true". This caused "let" assignments whose left- or right-hand sides were NaN to go awry. (One must work to compute NaNs, but assignments involving them should work.) Fix a bug with "write table"'s handling of symbolic parameters: pure numeric values were sometimes converted to decimal strings. Fix glitch in the builtin .tab handler in distinguishing '123' and 123 when reading symbolic parameters from columns after one or more skipped columns. Fix another glitch in the same handler that prevented reading Infinity and NaN in .tab files. Fix fault that arose when a symbolic imported function returned NULL. Turn such a return value into "". Addition to funcadd.h: char* getenv(const char*), which lets imported functions and table handlers access the current environment (as modified by option, environ, and problem commands). Only use getenv (i.e., (*ae->Getenv)) if al->ASLdate >= 20000608. 20000613 Fix bug in unloading libraries that could make calls enrolled with at_exit happen at the wrong time. The problem was revealed in an example involving a table handler, but could also afflict imported functions that call at_exit. The bug could bite either at the end of execution or during an "unload" or "reload" command. Fix a bug with the shell command that afflicted only Win32 binaries and was probably introduced with the changes of 19990804 to permit blanks in solver and directory names: shell commands with arguments had the arguments treated as part of the command name, resulting in failure of the shell command. 20000615 Fix longstanding bug with "close" command applied to a file used in a "read ... dig == 0 when al->deriv == 0. 20000719 Fix (rarely seen) bug in handling defined variables: if a nonlinear defined variable (v5 in the example below) had some linear terms and appeared only in the right-hand sides of linear defined variables (v7 below, which is linear after x[4] is fixed) used nonlinearly (in v8), incorrect function values could result. Example: var x{i in 0..4} := .1; var v5 = x[3] - x[1]*x[2]; var v6 = x[2] / x[0]; var v7 = v5 / x[4]; var v8 = v6 - x[1]*v7; minimize foo: x[2]*v8; s.t. fix_x4: x[4] = .099; In this example, the linear part of v5 was contributed twice to v7. Specifying "option linelim 0;" bypassed the bug. Fix glitch with expressions involving a simple or subscripted objective, appearing in commands (as opposed to declarations): if the objective involved one or more defined variables and nothing else forced the presolve state to be current, the objective value could be miscomputed. Example: var x; var y = x + 1; minimize zot: y^2; print zot; # printed 0 rather than 1 20000724 Fix longstanding bug with "reset data": rereading an indexed collection of sets sometimes led to memory corruption. 20000814 Fix bug with defined variable/bound sequences of the form var v1 = nonlinear expression; var v2 = const*v1 + const; s.t. v2_bound: lowerbound <= v2 <= upperbound; in which constraint v2_bound appeared to be lost. Example: var x := 1; var y = x^2 + 1; var z = y + 2; s.t. zot: z <= 5; minimize foo: (x + 4)^2; Work-around: option substout 4; Recognize command-line option -b? (and fix fault with -bx, where x is an invalid -b option). 20000825 Adjust prompts to indicate "Waiting for end of line after #" and "Saw /*; waiting for */". To indicate the former, # is inserted at the start of the prompt; for the latter, * is similarly inserted. If the prompt string would otherwise be the empty string, nothing is inserted. Fix bugs with redeclaration of an entity assigned values in a (previous) data sections when the redeclaration declares the same kind of entity: (1) require "reset data" or "update data" before accepting new or updated data for in a data section and (2) fix a bug that could cause memory corruption if enough enw subscript values were assigned in a subsequent data section. Example: set A; param p{A}; data; param :A: p := a 1 b 2 c 3; display p; redeclare param p{A} >= 0; update data A, p; data; param :A: p := a 1 b 2 c 3 d 4 e 5 f 6 g 7 h 8 i 9 j 10; display p; # fault or wrong values due to bug Change to builtin "sub" and "gsub" functions: treat \n as newline (as in the "sam" editor of plan 9). Thus the pattern (.|\n)* now matches everything until the end of the string, including intervening newlines. Fix bug with use of synonyms in loops: some expressions were not re-evaluated, such as 1.._ncons in the display command in the following example: model diet.mod; data diet.dat; for{j in NUTR} { display j, 1 .. _ncons; drop diet[j]; } Make variable.no independent of whether the variable is fixed. Adjust presolve to make more ambitious the extended presolve iterations that occur when $presolve > 1: when a variable is fixed, consider implications of modified constraints involving two or more remaining (unfixed) variables in the current extended presolve iteration rather than the next one; and when fixing two or more variables leads to simplifications of the same constraint, the first with implications for one bound (lower or upper), the second for the other, consider implications for both bounds. (Temporarily, option presolve_debug 3 suppresses these changes.) Have presolve use values of $presolve_eps, $presolve_fixeps, and $presolve_inteps that are rounded up by function R(x) = if x > 0 then 1.1^ceil(log(x)/log(1.1)) else 0. Encode the relevant values in .nl file headers and arrange for the solution command to issue option commands if necessary to set $presolve_eps, $presolve_fixeps, and $presolve_inteps to values that permit correctly reading the solution. In addition to using option presolve_fixeps during simple presolve reductions (those for $presolve = 1) to decide when tighter bounds lb and ub on a variable v (lb <= v <= ub) are close enough to consider v fixed (when ub - lb <= R($presolve_fixeps)), also use this test when extended presolve iterations ($presolve > 1) reduce ub - lb; in this case, v is fixed to if ub < 0 then ub else if lb > 0 then lb else 0. (Temporarily, "option presolve_debug 8" suppresses this.) New system params _presolve_eps_L and _presolve_eps_U indicate whether changes to $presolve_eps would matter: they should not matter if _presolve_eps_L <= R($presolve_eps) < _presolve_eps_U. Similarly, new system parameters _presolve_fixeps_L _presolve_fixeps_U _presolve_inteps_L _presolve_inteps_U give corresponding bounds for $presolve_fixeps and $presolve_inteps: presolve results should not change as long as _presolve_fixeps_L <= $presolve_fixeps < _presolve_fixeps_U and _presolve_inteps_L <= $presolve_inteps < _presolve_inteps_U. Rerun presolve if $constraint_drop_tol changes. This already happened for $presolve_eps, $presolve_fixeps and $presolve_inteps. Arrange for option show_stats 2 to produce the same output as option show_stats 1, plus an additional line about $presolve settings: report current $presolve and indicate whether a smaller value would suffice (albeit saving no time) or a larger value might give stronger bounds. In the latter case, the possible improvements are sometimes minuscule. Two new builtin params give values that help determine the new line: _presolve_req is the value of $presolve needed to reproduced the results of the last run of presolve. (This value has long been encoded in .nl files to facilitate correctly reading .sol files.) _presolve_sug = presolve_req if further presolve iterations would not help and is otherwise either $presolve + 1 (if $presolve > 0) or 10 (if $presolve = 0). Values of $presolve_*eps shown in output for $show_stats = 1 or 2 are now rounded to assure changes to R($presolve_*eps). 20000906 Fix bug in the error message "can't compute member(0,...)": the set name should have appeared. Fix a bug that gave rise to the error message "OPDIV botch in e2v". A simple example seems hard to find. Arrange for redeclarations of unsubscripted sets with no := clause to retain their values (when redeclared to be a set of the same arity). Example: set S ordered; data; set S := a b c; display S; redeclare set S ordered; display S; # OK now; previously "no data for set S" Some commands, such a (partial) fix command (involving a hitherto uninstantiated variable after some other command had instantiated a variable declared after the one being partially fixed), sometimes appeared to require instantiating the current problem, possibly leading to an error message about something that should not have mattered to the current command. Example: param p; var x{i in 1..3} := i + 4; var y; var z = (x[1]-p)^2 + 3; minimize zot: z^2 + (y-3)^2; let y := 5; # instantiates y, which was declared after x fix x[2]; # error processing var z: no value for p Fix a small memory leak. 20001002 Add "contains" to the reserved-word list. (Later, it will be a set-comparison operator, with A contains B <==> B within A.) Fix glitches in computing reduced costs (_var.rc) on problems involving piecewise-linear terms that can be linearized without use of integer variables (i.e., without special-ordered sets). Fix glitch in handling display _con.symbolic_suffix: adjustments for complementarity constraints were missing, leading, e.g., to possibly incorrect values of "display _con.status;". 20001006 Turn v1*sum{i in A} p[i]*v[i] into zero, where v1 and v[i] are variables and p[i] == 0 for all i. 20001009 Fix bugs with "unload" and "reload": 1. In a sequence of the form load foo.dll; solve; unload foo.dll; load foo1.dll; solve; in which foo.dll and foo1.dll define different functions of the same name, $ampl_funclibs was not updated for the second "solve", causing the solver to load foo.dll instead of foo1.dll. (If other changes caused presolve to run again, this bug did not bite.) 2. For all but the last-loaded library, "unload" might not have had the system actually unload the specified library, which would cause confusion if one unloaded a library, recompiled it, then loaded it again. 20001018 Fix bug in conveying complementarity conditions: a variable with two finite bounds was sometimes paired with a one-sided constraint, possibly leading solvers to an incorrect interpretation of complementarity conditions. Example: var y; var l{1..2} >= 0; s.t. c1: 0 <= y + 10 complements l[1] >= 0; s.t. c2: 0 <= -y + 20 complements l[2] >= 0; # -20 <= y <= 10 was incorrectly paired with l[1] >= 0 # Instead, l[1] >= 0 should have been paired with y >= -10 20001025 Fix bug with complementarity constraints: when presolve resolved a complementarity condition, changing an inequality constraint to an equality, if the constraint involved two or more variables with infinite bounds and coefficients of suitable signs, presolve could incorrectly deduce a stronger bound on one of the variables, leading possibly to transmitting an incorrect problem to solvers. Example of incorrect bound deduction (var.lb2 and var.ub2 are the strongest deduced bounds): var x{2..6} >= 0; s.t. c1: 4*x[4] + 4 <= x[2] + 3*x[3] complements x[6] >= 0; s.t. c2: x[5] + 5 <= 4*x[6] complements x[4] >= 0; display _svarname, _svar.lb2, _svar.ub2; # x[3].lb2 botched: should be 0 but was 1.33333 Arrange for "drop foo;", "restore foo;", "fix goo;" and "unfix goo;" not to require generating foo or goo (whereas if, say, foo is subscripted, as in "drop foo[3];", then generating foo is unavoidable and all data required to instantiate foo must then be present). This was only an issue after references to a subscripted named problem whose indexing set involved other declared entities. Example: param p; set I := 1..2; problem zork{I}; var x{1..p}; var y; minimize zap: y^2 + sum{i in 1..p} x[i]^2; problem zork[1]; drop zap; Fix a bug in recovering from missing data in contexts where a new variable was to have been (re-)instantiated that was declared before a previously instantiated variable: if nothing subsequently caused a variable to be (re-)instantiated, confusion was possible -- misnumbered variables or an erroneous "no variables used" message. Example: param p; set I := 1..2; problem zork{I}; var x{1..p}; var y; minimize zap: y^2 + sum{i in 1..p} x[i]^2; minimize zip: sin(y); problem zork[1]; fix x; problem zork[2]; drop zap; solve; # OK problem zork[1]; solve; # error generating x: no value for p problem zork[2]; solve; # erroneous "No variables used." 20001030 Fix bug with complementarity conditions: if (perhaps after some presolve reductions) one of the constraints in a complementarity condition was a simple-bound constraint implying that the variable in question could be fixed, the complementarity condition was not treated as resolved, leading to a mysterious error message about "presolve has k = ...". Example: var z{1..2} >= 0; s.t. c: z[1] <= 0 complements z[2] <= 0; solve; # presolve has k = 2, P.nfc = 0 20001109 Fix fault with some references to unavailable functions. Example: function zork; var x; minimize f: zork(3)*x; solve; # fault when zork is not in amplfunc.dll 20001117 Fix a bug with the printf command, sprintf function, and formatted pipe functions: results (in the case of pipe functions, the formatted string being sent to the function) over 256 bytes long were sometimes truncated. An example where the bug bit: printf "%s\n%s\n%s\n%s\n%s\n%s\n%s\n%s\n%s\n", {i in 1..9} ('Line ' & i & ' of a very long string intended to illustrate a printf bug'); Fix bug in handling functions of more than 8 arguments on machines with 64-bit addressing. The bug also afflicted imported functions with more than 2 explicit domains. 20001128 Fix bug writing tables of zero rows (which were inadvertently described as having one row, sometimes leading to a fault). Fix obscure bug on machines with 64-bit addresses in reading .bit files (builtin binary table files) written on machines with other endian-ness (i.e., written on a big-endian machine and read on a little-endian machine or vice versa). ampltabl.dll: handle .dsn files better (no longer require them to contain DBQ= lines); no longer require DSN=... strings to contain DBQ=...; permit DRIVER=... to specify a full connection string (as well as DSN=...); handle driver-specific types in addition to the ODBC standard ones. 20001129 Fix bug simplifying nonlinear "if" expressions. Sometimes the result of comparison operators was erroneously regarded as known; since the bug was due to initialize a variable in this case, behavior was system dependent. 20001130 Fix bug introduced 20000825 in reading .sol files that could lead to inappropriate option changes (with the relevant option command(s) printed during solution reading) for $presolve_fixeps, $presolve_inteps, or $presolve_eps. If the bug bit (an apparently rare event), the solution might be read incorrectly. 20001206 Arrange for reset; print 'stuff' >zap; reset; to close file zap. 20001228 Fix bugs (faults) with "fix all;" and "fix all[3];", and fix a glitch that added confusion to the error message that the latter now elicits (about things that must or cannot be subscripted). 20010102 Fix a bug with let S := {}; within loops. At the second or subsequent loop execution, it was possible for S to be misassigned, which led to a fault in the complicated example that revealed this bug. 20010123 Arrange for reset; print Uniform01(); reset; print Uniform01(); to print the same value twice. Previously the random number sequence was only reset if at least one declaration had appeared before "reset;". Fix bug that could give error message "unexpected type 0x87 in e2v()" when constraints or objectives involved the "mod" operator. 20010129 Fix error-message handling bug that caused set I := 1..3; var x{I} >= 0; param a{I} default Infinity; s.t. bletch: sum{i in I} a[i]*x[i] = 1; data; param a := 1 1.2, 3 4.5; write 0; to elicit error message tva == 0 in mgwhere() Now it gets Error executing "write" command (file goo, line 4, offset 130): error processing constraint bletch: can't multiply x[2] by Infinity Fix a glitch in handling {if logical_expression} in the ... of <<...; ...>>, as described on p. 312 of the AMPL book. For example, param p default 3; var x; minimize zap: <<1,{if p in interval[0,2.9]} 2, 3; -1,{if p in interval[0,2.9]} p, 4,6 >> x; provoked an erroneous complaint about "iteration over unordered set". In iterated commands and expressions, permit {if logical_expression} wherever indexing expressions were previously allowed. 20010130 Fix another bug that led to "tva == 0 in mgwhere()". Example: set A; set B{A}; set C := intersect{i in A} B[i]; data; set A := ; display C; For completeness, permit {if logical_expression} as the indexing in iterated reduction operators (sum, prod, min, max, exists, forall). 20010208 Fix possible fault with handling identical slices in the same indexing set. The fault could occur when a command involving the index set was executed twice in succession for the same slice. Example: set A dimen 2; data; set A := a 1, a 2, a 3, b 2, b 3; for{k in 1..3} for{(i,2) in A, (j,2) in A} print i,j,k; (The fault did not occur with "(j,2)" changed to "(j,k)", since the slices where then different on successive executions of the inner "for" loop.) Fix bug (infinite loop) in "print 2.47032822920623272e-324;". Fix bug in handling different slices involving the same expressions over a subscripted set. (The fact that the expressions differed was overlooked in this particular context.) Example: set A{1..2} within 1..3 cross 1..3; param p {k in 1..2, i in 1..3} := sum{(i,j) in A[k]} j + 100*sum{(j,i) in A[k]} j; # Param p involves two different slices over A[k] in the same # context and involving the same expression, in this case "i". data; set A[1] := 1 2 1 3; set A[2] := 2 3 3 1; display p; # gave all zeros 20010209 Fix parsing glitch with single-step mode: when a stand-alone "if" command (not within a compound command) was not followed by an "else" clause, an erroneous "syntax error" message was possible after execution of the "if" command. Work-around: add an extra semicolon. Example: param p default 1; option single_step 1; if p == 1 then let p := 2; #extra ; here = work-around printf "p = %.g\n", p; # got "syntax error" 20010212 Fix bug (fault) handling certain set expressions, such as slices of subscripted sets, in recursive param and set definitions. The bug appears to have been around since at least 19971218. Example: script on p. 314 of the AMPL book plus some simple data: set nodes; set arcs within nodes cross nodes; param max_iter := card(nodes)-1; set step {s in 1..max_iter} dimen 2 := if s == 1 then arcs else step[s-1] union setof {k in nodes, (i,k) in step[s-1], (k,j) in step[s-1]} (i,j); set reach := step[max_iter]; data; set nodes := a b c d e f g h; set arcs := a b a c b d c e f g f h; display step, reach; This sort of thing is more efficiently done with commands (which also avoids the bug): set nodes; set arcs within nodes cross nodes; param t; set reach default arcs; data; set nodes := a b c d e f g h; set arcs := a b a c b d c e f g f h; repeat { let t := card(reach); let reach := reach union setof{k in nodes, (i,k) in reach, (k,j) in reach} (i,j); } while(card(reach) > t); display reach; 20010215 Fix bug introduced 20010102 in handling named (non-subscripted) sets that happen to have the same values as other sets, but then are assigned different values: the new values appeared to be lost. Example: set A; set B; data; set A := a b; set B := a b; display A, B; let B := {'c','d'}; display B; # a b 20010226 Fix a possible fault or unpredictable behavior with "write b;" (which should write ".nl"). 20010227 Fix a glitch in the Win32 ampl.exe when the current directory is on certain 8.3 file systems: due to changes on 20010226, "solve;" might elicit an error message about an unwritable file. 20010307 Fix a rarely seen reference-counting bug that could cause the value of a seemingly random set to change. The bug could only bite under the "right" conditions, which seem hard to capture in a simple example. 20010312 Fix a glitch that could cause "" to appear as the name for some generated constraints or variables in, e.g., "display _sconname;" when constraints were added in the course of linearizing nonconvex piecewise-linear terms. Fix bugs in handling very long character strings and single lines with no white space between tokens. For (ephemeral) constraints and variables introduced to linearize nonconvex piecewise-linear terms, fix bugs in assignments to _svar and quietly forbid assignments to suffixes (which do not exist in a general sense for these ephemeral entities). 20010316 Fix bug introduced 20010307 with "reset data;" or "reset data p;" after a value has been assigned (via data section or "let") to an unsubscripted symbolic parameter. Example: param p symbolic; data; param p := 3; print p; reset data; # fault 20010322 Again adjust "option randseed '';" and AMPL's -s command-line option to make it more likely that successive invocations will give different values. The changes are visible mainly on Microsoft systems. Eliminate possible faults after a "reset;" while single stepping or after sequences of the form option log_file 'something'; # declare something and do some stuff reset; option log_file ''; # subsequent trouble was possible /netlib/ampl/student/mswin/sw.exe.gz: make complaints about an "Edit-control bug" less likely (under NT -- W98 has a different bug). 20010328 Fix a bug (fault) with string-valued function references in recursive parameter definitions. Fix a bug with "sub" and "gsub". Memory was sometimes corrupted. 20010405 New Caution: unless $Caution is 0, warn about numeric option settings where part of the option string is ignored. In response to "solve;" after a previous command has caused presolve to determined the solution or eliminate all variables, if $solver_msg is nonzero (as it is by default), print solve_message again -- it should also have appeared after the previous command, and it explains the situation. Example: var x >= 1 <= 2 := 3; var y = x + 2; solve; # "No variables used after elimination of defined variables." solve; # now gives the same message, rather than silence. 20010412 Fix a glitch parsing multi-line character strings that end with an even number of backslashes (\). Adjust "read table" and outarg processing so (as with "let" and data sections) values for subscripts for which heretofore there was no value do not cause recomputations of things that depend on the param being updated. When reading stdin in interactive mode, completely clear the error indicator when not in the middle of a statement. Previously the first command after a declaration error was suppressed. 20010417 Make the "Too much memory used" message print a positive number when the number is >= 2^31. Fix bug introduced 20001006 in simplifying the product of two apparently nonlinear expressions. When the right-hand "nonlinear" expression turned out to be a constant, the constant was erroneously assumed to be 0. Example: var x; param t default 0; minimize zap: (1 - x*exp(t*x))^2; solve; # said "No variables used." 20010507 Fix bug in handling uses of nonlinear objective values in commands: some sequences of commands (instantiating variables in the objectives but not the objectives themselves) could fault. Example: var x; let x := 5; minimize zot: (x-1)^2; if zot >= 3 then print 'yes'; # faulted (but "zot.val" was OK) In imported (user-defined) functions, have printf("%s",0) print "" rather than fault. When presolve determines a feasible solution, set solve_result_num = 99. 20010514 Fix a glitch with "unused" variables -- ones that do not appear in the presolved problem. Under somewhat unusual conditions, "unused" variables that appeared in constraints deduced always to be slack were not projected onto their bounds, which could result in infeasible constraints after the message "Solution determined by presolve." Example: var x{i in 0..3} >= i <= i^2; s.t. zot{i in 0..2}: x[i] <= x[i+1] + 1; solve; # Solution determined by presolve display x, x.slack, zot.slack; # x[3] was being left 0 20010517 Fix a bug that appears to have crept in at version 19980127: an incorrect opcode was emitted for character-valued "if" expressions used as arguments to imported functions, as in var x; function foo; minimize zot: foo(x, if x >= 3 then 'abc' else 'wxyz'); Fix a bug computing dual values in some problems involving both linear and nonlinear defined variables used in certain ways: it was possible for some dual values to be miscomputed. Example: set I := 1..3; var x{I}; var y = sum{i in I} (x[i] - i)^2; var z = sum{i in I} i*x[i]; var w = y + z; minimize zot: w; s.t. convex: sum{i in I} x[i] == 1; s.t. nneg{i in I}: x[i] >= 0; solve; display nneg; # nneg[1] should be 0.5, not 0. 20010524 Fix a bug in handling complementarity problems involving defined variables that may be split into linear and purely nonlinear parts. 20010613 Fix bug (fault) with the following erroneous input: data; 'stuff' Under -R, restrict remove commands to alphanumeric file names (and depend on directory and file permissions to control what can and cannot be removed and written). Adjust treatment (probably introduced 19941003) of infeasible problems: after presolve detects infeasibility, whether a second solve or write command is allowed to proceed is now controlled by new option infeas_clear, whose default value 1 allows the commands to proceed when issued in interactive mode on stdin. Specifying "option infeas_clear 2;" restores the former behavior of letting second solve and write commands proceed under all circumstances, and option infeas_clear 0 treats a second such command just like the first. 20010709 Fix bugs (possible faults, or worse) with the delete command applied to variables, constraints, and objectives under the "right" conditions. 20010724 Fix longstanding bug with display commands iterated over sets of arity > 1: a fault or a surprising error message about too much memory (or worse) was possible. Example: set A; set B dimen 2; param p{A, B}; data; set A := a b; set B := (w,x) (w,y) (x,y) (x,z); param p [a,*,*]: x y z := w 1.1 2.2 . x . 2 3.3 [b,*,*]: x y z := w 4.4 5.5 . x . 4 6.6; display{(j,k) in B}: sum{i in A} p[i,j,k]; 20010816 Fix bug in handling table declarations with one key column that coincides with an intermediate string in the list of strings that precedes the key-spec: table handlers saw an incorrect list of strings. In some cases, this could result in a fault, as in set E; table Zap IN 'ODBC' 'foo.xls' 'Blop' 'verbose': E <- [Blop]; read table Zap; display E; Fix bug with "option substout 1": constraints of the form x = x + stuff_not_involving_x led to the error message "solve_out bug!". Example: var x; var y := 1; minimize foo: sin(x); s.t. zot: x = x + y^2 - 2; option substout 1; solve; display x, y; # solve_out bug! Sample ampltabl.dll for MS Windows: when reading tables in an explicitly specified file, use the full pathname in "DBQ=..." to avoid confusion with similar names known to the ODBC Data Source Administrator. When writing tables, the sample ampltabl.dll still tries to use "DSN=..." if possible. Note that to permit writing .xls files, it's necessary for the ODBC Data Source Administrator to have a DSN (e.g., a User DSN) associated with the Microsoft Excel driver with the "Read Only" box unchecked (under Options in ODBC Microsoft Excel Setup). 20010817 Fix glitch with update data for subscripted symbolic parameters: all subscripts were treated as the first subscript. Example: set A; param p{A} symbolic; data; param :A: p := a abc b def c ghi z xyz ; display p; update data p; data; param p := c pqr b jkl; display p; # wrong p['a'], p['c'], and p['b'] Again adjust things so reset; print Uniform01(); reset; print Uniform01(); prints the same value twice -- the previous attempt (20010123) was wrong. This change only matters if there were no declarations before the "reset;". 20010927 Fix an obscure presolve bug that led to an error message of the form presolve has k = 708, P.nfc = 704 20011004 Fix a glitch in handling /* ... */ comments: under unusual conditions, an inappropriate end-of-file message was possible. Fix a bug in handling suffix declarations with complicated bounds on the suffix values: a fault was possible. 20011119 Fix glitches with printf: printf "%3.0f\n", .1; # gave " 0" rather than " 0" printf "%3.-2f\n", 1; # gave " 00" rather than " 0" Ensure that A union B puts members of A first. Usually it did, but under some conditions the members of B came first. Fix bug with "reset data x;" when x is a recursively defined param or set: entities that depended on x were sometimes not recomputed. Example: param p{i in 1..6} := floor(Uniform(1,5)) + if i > 1 then p[i-1]; set S := setof{i in 1..6} p[i]; display S; reset data p; display p; # S was recomputed correctly without this line. display S; # retained the old values (after "display p;") 20011206 Fix bug with "solve": with a sequence of commands that first cause the current problem to be instantiated, then change the values of some variables, then have a "write" command followed by a "solve" command, the "solve" command wrote a new (temporary) .nl file rather than using the one written by the "write" command. One way this manifests itself is that the "write" command writes auxiliary files based on $auxfiles, whereas when the "solve" command writes a temporary file, it writes auxiliary files based on $($solver)_auxfiles. Example (using solver linrc from "Hooking Your Solver to AMPL"): model diet.mod; data diet.dat; write 0; #instantiate the problem let Buy['BEEF'] := 1; option auxfiles rc, solver linrc; write gdiet; solve; # names from diet.mod were not shown Fix a bug (possible fault, or worse) in simplifying nonlinear expressions involving piecewise-linear terms when presolve fixes a variable. (Such expressions are best avoided when using solvers that expect continuous derivatives.) Adjust processing of (the default) "option linelim 1" so defined variables that are not involved in the current problem will not be affected by "option linelim 1" (which took considerable time in a motivating example). The temporary interpretation of the "4" bit of $linelim, introduced 20000327, is rescinded. Now, temporarily (for debugging), when the "4" bit of $linelim is on, today's change to processing for "option linelim 1" will be suppressed. 20011211 Fix a bug with the "option linelim 1" processing change of 5 days ago (version 20011206) that could lead to a fault or wrong results. 20011213 Fix fault in parsing { ... repeat while(...){...}} (with no semi-colon after the first right-brace, the one meant to end the repeat loop without the optional trailing while or until clause). Infer the missing semi-colon. Example: param p default 0; for{i in 1..6} { repeat while(p < 3) { display p; let p := p + 1; } # missing ; here } Fix longstanding bug in handling a nonlinear piecewise-linear expression applied to a subscripted variable, where the subscript only involves dummies from an outer context and the expression appears in an inner one (such as a sum). The bug led to a surprising error message ("tva top error") in the following example: set I; set K; var b{I} binary; var x{K}; param np integer > 1; set J ordered := 1..np; set J1 ordered := 1 .. np-1; param p{J1}; param q{J}; var d{k in K} = sum{i in I} b[i]* <<{j in J1}p[j]; {j in J}q[j]>>x[k]; minimize zap: sum{k in K} sin(d[k]); data; set I := a b; set K := c d; param np := 2; param : p q := 1 1 2 2 . 3; write 0; # gave "tva top error" 20011231 Fix bug with single-step mode "next" command: a fault or surprising error message was possible, e.g., with "next" applied to the repeat loop in NEW/LOOP1/steelT.sa6 in the AMPL web site. New option presolve_assoc (default 7) affects simplification of constants in nonlinear contexts during presolve: sum of 1 ==> permit using associative law on + and - operations; 2 ==> permit using associative law on * and / operations; 4 ==> permit using distributive law on const*(const*thing +- const*thing). Change command-line option -o to write a .nl file at the end of the last command-line input file if a "solve;" command would write a new .nl file. (Note that if no command-line input files are given explicitly, then stdin is treated as the single command-line input file.) 20020129 Fix a bug (possible fault) simplifying certain complicated nonlinear expressions involving constant terms. A simple example seems hard to find. Fix a bug that led to the surprising error message unexpected nonvariable type 5698 = 0x1642 in eput in a complicated example. 20020409 Fix a bug simplifying complementarity conditions. Example: var x; var y; s.t. foo: x <= 1 complements y >= 0; s.t. ylb: y >= 1; write 0; got error message "presolve has k = 0, P.nfc = -1". Fix a bug with { cmdlist } at the outermost level. A fault (or worse) was possible after a subsequent error message. Fix a bug with "reset data foo;" when foo had a := value involving random functions: entities that depended indirectly on foo were not necessarily updated automatically, e.g., at subsequent solve or write commands. Give a less confusing error message for attempts to assign symbolic suffix values when the corresponding _table option has the wrong form. Warn about attempts to delete system entities (such as synonyms and system sets), rather than silently ignoring delete commands that mention them. Warn of inappropriate use of suffixes on synonyms. Fix glitch in display of entity.suffix_num: sometimes "_num" was omitted. Suppose $send_suffixes is 0 or 1. Then solvers are (and have always been) instructed not to return any suffix values. Now if the solver returns suffix values anyway or a "solution" command sees updated suffix values, the returned suffix values are ignored. For properly behaving solvers linked with interface library versions >= 20020402, adjust the reading of .sol files to omit the backspaces that used to appear when the solvers did not report any option settings. 20020424 Fix bug with "let{...} S := set_expr;" when S is a set and the set_expr involves S: parts of the set_expr were not recomputed. (It's generally better to use other constructs, such as setof{...} or union{...}, but iterated let commands should work on sets.) Fix bug handling defined variables when $presolve is 0: empty constraints were mishandled, leading perhaps to a fault (or worse). 20020425 Fix bug introduced 20020424 that affected "let S := set_expr" with set_expr involving S and the let command contained in a "for" loop. During the second and subsequent loop iterations, set_expr was miscomputed. 20020426 Refix bug introduced 20020424 that affected "let S := set_expr" with set_expr involving S and the let command contained in a "for" loop. Last night's fix did not work in all cases. 20020503 Complain about "model ..." within compound commands or during single stepping. Previously, for{i in S} { model (i & '.mod'); } could fault. Add SnprintF and VsnprintF to AmplExports, to make snprintf and vsnprintf available to imported functions. 20020508 To avoid breaking some scripts, again permit "model (...);" within compound commands when ... does not involve loop dummies. Note that the effect is similar to "model; include filename" in that the file is read before the command is executed. In particular, param p symbolic default 'zip'; for{i in 1 .. 2} { let p := 'zap'; model (p); } will read file 'zip' once -- and not read 'zap' at all. 20020516 Fix a memory-corrupting bug with sequences of commands that solve a problem with nonconvex piecewise-linear terms, update something (without "reset;"), and solve another such problem. 20020528 Fix an obscure bug that caused a fault when encountered (in accessing an "unused" defined variable). A simple example seems hard to find. (Inexact date; change made after discussion in May 2002. This note added 20210223.) Allow "= expr" to act as ":= expr" in param declarations when new option old_param_eq has its default value 0. This is for declaring a derived parameter whose value is computed from the given expression "expr". Specifying "option old_param_eq 1;" will cause "= expr" to have its old meaning, which was to specify a check on the param value. Option old_param_eq also affects the "show" command, which prints "= expr" in a relevant param declaration when $old_param_eq has its default value and ":= expr" when $oldparam_eq is 1. 20020602 Fix possible fault (or worse) in recovering from an error that terminates a "read" command. Fix a bug in read commands: some expressions involving loop dummies were not recomputed. For example, in param n := 3; param foo{1..n}; param goo{i in 1..n, j in 1..foo[i]}; data; param foo := 1 4 2 3 3 7; read {i in 1..n} ({j in 1..foo[i]} goo[i,j]); 10 20 30 40 1 2 3 100 200 300 400 500 600 700 display goo; foo[i] was not recomputed when loop dummy i changed. This resulted in the surprising error message "invalid subscript goo[2,4]". 20020622 Fix some minor storage leaks that become visible with many repetitions of a "commands" command. Workaround: use "include" when the commands file does not change. Suppress the complaint about a scalar variable or parameter not taking any subscripts when it is iterated as a constant output column in a table. Workaround: add 0 (or, for symbolic param, concatenate ""). Fix long-standing reference-counting bug in turning {..., {A,B}, ...} into {..., A,B, ...}. Workaround: write the simpler forms yourself. 20020708 Change to handling of multiple objectives: previously objectives were always reordered so that nonlinear objectives came first. Now the default is not to do this reordering. Option nl_permute is extended to allow indicating that objectives should be reordered as heretofore. $nl_permute is now the sum of 1 ==> reorder constraints 2 ==> reorder variables 4 ==> reorder objectives The default value for $nl_permute remains 3. Fix bug handling multiple objectives: when objectives were reordered because some nonlinear objectives were declared before linear ones, the sense of optimization (minimize or maximize) was not similarly reordered. Obscure correction: in block mode, error messages from "read table" and "write table" will now have "kind" values "read_table" and "write_table" rather than "read table" and "write table". 20020716 Obscure enhancement to handling of complementarity constraints of the form equation complements unbounded_variable: indicate to solvers that the unbounded variable is associated with the equation. Turning the "4" bit of $compl_warn on (usually by "option compl_warn 5;") suppresses this adjustment. Fix bug in generating variable names after redeclare under complicated conditions. Example: set I := 1..2; set J := 1..5; param A{I,J} := Uniform01(); var x{J} >= 0; var u{I}; dual{j in J}: sum{i in I} A[i,j]*u[i] <= 1; con{i in I}: sum{j in J} A[i,j]*x[j] == 1; problem p: x, u, dual, con; problem p; redeclare s.t. dual{j in J} : sum{i in I} A[i,j]*u[i] <= 1; expand _scon[1]; # showed x rather than u as variables Fix bug in "display _ccon;", which should give the same results as "display _ccon.slack;". Fix obscure bug with iterated "let" commands involving an entity with a constant subscript appearing on both sides of the := operator: expressions involving it on the right-hand side were not recomputed (as it was erroneously regarded as constant). Example: set A; param p{A} default 4; data; set A := a b c; let{i in 1..10} p['b'] := p['b'] + 1; display p; # showed p['b'] = 5 rather than 14 Fix similar bugs involving synonyms and declared suffixes. Add #ifdef __APPLE__ stuff to funcadd1.c and mac.c for Mac OS X. 20020801 Fix bug with declared suffixes on subscripted problems (an obscure feature): a sequence of the form set A; problem foo{A}; suffix zot; problem foo[something]; let foo[something].zot := ...; problem foo[something_else]; let foo[something_else].zot := ...; could have corrupted memory. Tweak to single-step mode: accept a semicolon or white space after the optional repetition count after "next", "skip", and "step". Fix glitches with the display command's handling of $display_width when formatting 2D tables: sometimes a column heading $n appeared unnecessarily (and "$n = ..." lines sometimes incorrectly involved quotes). Fix bug writing .nl files for complementarity problems when $var_bounds is 2: incorrect complementarity conditions might have been indicated. In problems with variables in singleton complementarity constraints, let the complementarity constraints imply the variable bounds when $var_bounds is 1 (the default), unless the variables are also matched with other complementarity conditions. Example: var x; var y; s.t. c1: x >= 0 complements y >= 0; s.t. c2: x + y = 1; display _varname, _var.lb, _var.ub; # above now shows no bounds on y 20020806 Fix a bug (possible fault or worse) in a sequence of the form solve; # with defined variables that get split # into linear and nonlinear parts # ... solve; # with no defined variables Fix bugs (faults) in handling table declarations involving suffixes on subscripted entities appearing within expressions, such as diet[i].slack and _con[i].slack in the "such that" parts of the set expressions in the following table declarations: model diet.mod; table ds OUT 'ds.tab': [NUTR] {i in NUTR: diet[i].slack > 0} diet[i].slack ~ Slack; table dsyn OUT 'dsyn.tab': [I] {i in 1.._ncons: _con[i].slack > 0} _con[i].slack ~ Slack; 20020807 Fix bug in computing cc1.Lbody and cc2.Rbody, where cc1 and cc2 are complementarity constraints with a single expression (no finite bounds) to the right and left, respectively, of "complements": the constant term (if any) was omitted. (This did not affect the problem seen by the solver.) Fix bug with "expand _con[n];" when _con[n] is a complementarity constraint: for n > 0, the wrong constraint was expanded. 20020814 Fix faults with drop {...} X; where X is an indexed collection of objectives or constraints; this includes {if logical_expression} indexing. Fix analogous faults with restore, fix, and unfix. In the display command's printing of 2D tables, permit the final ":=" part of the ": ... :=" header to exceed $display_width (to make the output a bit more regular). We can always recant this tweak if it turns out to cause more trouble than it avoids. Correct error message for set S ordered; data; set S := a b c d; reset data S; data; set S := a b '3' d '5'; print next('5',S); -- the set name "S" got lost in "reset data S", and the error message showed the entire contents of S. Now it just mentions "S". In error messages for next() and prev(), quote symbolic members (to distinguish, e.g., 3 and "3"). Fix bug with recursive set and parameter definitions involving certain complicated indexing expressions, resulting in a surprising error message: "cp_lookup: unexpected hit". Example: param N := 100; param odd {i in 1..N} := if i = 1 then 3 else min {j in odd[i-1]+2..odd[i-1]*2 by 2: not exists {k in 1..i-1} j mod odd[k] = 0} j; display odd; In set and param declarations, treat = (or ==) as a synonym for :=, so = behaves as it does in var declarations. For param declarations, this is simply an extension. For param declarations, it changes the meaning of =, which previously implied a surprising test that the param had the specified value. (We believe that = was almost never seriously used in this way, so that this change will be harmless. If not, "option old_param eq 1;" restores the old meaning of = in param declarations, at least for now.) Arrange that failed consistency checks detected during a "solve" command will result in solve_result = '?' and solve_result_num = -1, just as they are at the start of execution. Previously they retained their values from the last successful "solve". Change "s.t." to "subject to" in "show" and "expand" commands. 20020819 Fix some glitches (resulting in surprising error messages) with the handling of OUT and INOUT arguments in imported functions with certain signatures, such as (INOUT ...) and (IN, {i in A} (INOUT), OUT). Also fix some glitches in "redeclare function ...". Fix a bug with switching problems after an explicit environ command: the change to the environment was sometimes not recorded. 20020820 Fix bug in handling defined variables that only appear in piecewise-linear terms: they were treated as unused in a way that led to surprising appearance of nonlinearity. Temporary work-around: "option linelim 5;". Correct a botch in yesterday's saving of problem environments. tables/ampltabl.dll: report more when SQLNumResultCols fails unexpectedly. 20020823 Fix a bug in computing dual values for problems with nonconvex piecewise-linear terms. Reduced costs for variables appearing in such terms may not have been right, and a fault was likely if the problem involved defined variables. Fix a glitch with displays of scalars whose descriptions take more than one line: only consider the length of the last line when deciding whether $display_width permits showing the value on the same line. 20020827 Obscure bug fix to printf.c, fprintf, sprintf: on systems with IEEE arithmetic, get an explicitly requested sign of zero right under format %+e. Example: printf "%+e\n", -0; 20020917 Fix bug in parsing unquoted file names that contain nonalphanumeric characters (such as /, \, or :) and fall on a block boundary (multiple of 4096). Workaround: quote such file names. 20020924 Fix bugs with problem command: if foo is a subscripted problem, and "problem foo;" reports foo[subscript] current, have "problem foo;" and "delete problem foo;" discard the subscript. Fix fault with "problem foo{1..3}; problem goo;". When foo is a subscripted problem whose declaration does not specify a subscripted environment, have "problem foo;" make foo's environment current. Fix a rarely seen bug in presolve's conversion of nonlinear expressions to linear expressions (after fixing of some variables). If the bug bit, it could cause a surprising "vT bug" error message (or could cause some range estimates in presolve to be overly pessimistic, which might lead to worse presolve results but was otherwise harmless). Fix bugs handling double inequalities involving string expressions. Faults were highly likely. 20021002 Fix bug handling equality complements expression under option presolve 0. Workarounds: leave $presolve alone or simply express such constraints as ordinary equality constraints. Fix bug (fault) introduced 20020807 in computing cc.Lbody and cc.Rbody when cc is a complementarity constraint with a single inequality constraint on each side of "complements". Fix bug introduced 20020807 in computing .slack for complementarity constraints with a double inequality complementing an expression. Fix bug with "expand _scon[n];" when the declared constraint is not subscripted. 20021009 Fix an optimization bug with "first" and "last", which could be erroneously evaluated on an empty set when lifted to an outer context (in a complicated indexing expression). If the bug bit, it resulted in a surprising error message. 20021025 Fix bug that gave error message "OPDIV botch" when presolve detected division by zero. Example: var x >= 0; var y; minimize zot: x/y; s.t. ydef: y = 0; solve; Fix an obscure bug with "option substout 1;": after AMPL obtained a solution from the solver, it did not compute the values of variables eliminated from the problem sent to the solver until needed. Changes affecting or eliminating the defining constraints would then influence the values computed for such variables. Now their values are computed when a "solve" or "solution" command obtains the solution. Unlike variables declared with a defining value (= expression), a variable eliminated by "option substout 1" has never been automatically updated when something in its defining constraint changes. Arrange for variables to which piecewise-linear terms are applied to be defined by the linearizations of the terms. On problem 14.7(d) in the AMPL book (first edition; 17.7(d) of the second), this reduces the numbers of variables and constraints by 1/3. Turning the "4" bit of $pl_linearize on, e.g., with option pl_linearize 5; suppresses this change. Infer bounds of 0 and 1 on binary variables, even when they are relaxed by assignment of 1 to their .relax suffix. Fix a bug in processing complementarity constraints that caused inequality complements equality simply to be treated as the equality. Now an error message results. (The bug also let some inappropriate operators, such as <, >, and !=, slip by, which could have led to a subsequent fault.) 20021031 Fix another bug (possible fault or wrong values) in computing dual values for problems with piecewise-linear terms. 20021108 Fix a bug with use of at_exit() or at_reset() in imported functions: the library was freed before the functions registered with at_exit or at_reset were called, sometimes leading to a fault. 20021202 Fix glitch introduced 20021025 with variables involved in piecewise-linear terms: after a "solve" or "solution" command, a change (e.g. by a "let" command) that caused recomputation of defined variables could result in variables that appear piecewise-linear terms to have the wrong value. Using "option pl_linearize 5;" is a temporary work-around. Fix a performance glitch with "let" that caused an assignment to a numeric param of its current value to be treated as a change to the param's value. 20030103 Fix a bug (possibly leading to a fault or incorrect execution) in turning {..., A cross B, ...} into {..., A, B, ...} in sets and indexing expressions. Work-around: write the latter form. 20030204 Cut over to a version that works as described in the second edition of the AMPL book, including such relatively obscure features as new tabular forms in data sections, handling of "in union_of_intervals" phrases in variable declarations, and $(expr) string expressions. Some new constraint-logic programming features, described in INFORMS J. Computing 14#4 (2002) pp. 322-344, are also recognized. (Variables in subscripts are still missing.) 20030217 Fix a bug (introduced 20030204) in the "expand" command's handling of complementarity constraints: the wrong constraints were sometimes shown. 20030227 New option strict_ineq_warn (default 1) determines how to handle constraints involving a strict inequality when the constraint would be an algebraic constraint if the comparison were changed to permit equality (i.e., < were changed to <= or > were changed to >=): 0 ==> quietly treat the constraint as a logical constraint; 1 ==> print a caution and treat as a logical constraint; 2 ==> print a warning and reject the constraint. Before 20030204, AMPL behaved as though $strict_ineq_warn were 2. 20030310 Fix a glitch with random number generation in the Linux version of AMPL (introduced 20030204 by a compiler update): e.g., Uniform01() was often outside [0, 1]. 20030319 Fix glitches with "unload" (which did not completely unload the indicated library) and "delete function_name" (e.g., when the function name coincided with a builtin random function). 20030328 Fix a fault introduced 20030319 in handling ampltabl.dll after a reset. Fix a glitch that sometimes caused "unaligned access" errors on systems with 64-bit addressing. 20030331 Fix bugs (faults) with references to _svar[n].suffix when _svar[n] was created by linearization of nonconvex piecewise-linear terms. Adjust option relax_integrality so binary variables added to linear nonconvex piecewise-linear terms are retained under option relax_integrality 1; and are relaxed, with no suffix information about them transmitted to the .nl file, under option relax_integrality 2; Before this change, solvers might fault unless explicitly told to ignore SOS information when $relax_integrality was 1 and the problem had nonconvex piecewise-linear terms. 20030414 Fix bug introduced 20030204 in "option var_bounds 2": the count of Jacobian nonzeros was misreported, which could cause some solvers to fault. 20030502 Fix a glitch in reading .sol files: solvers are supposed to be able to assign suffix values to the current named problem, but such values were being ignored. 20030508 Fix glitches with some expressions involving _con, _scon, _obj, and _sobj when the relevant constraint or objective was partially dropped. E.g., "let _con[...] := ...;" and "print _obj[...];" were affected. Fix a botch (introduced 20030204) in the show command: it reported the wrong entity type for some things: "show e;" said "setvars: ..." instead of "environments: ...", "show su;" said "xhatBUGs: ..." instead of "suffixes: ...", and "show t;" said "s: ..." instead of "tables: ...". 20030513 Fix a longstanding and apparently rarely visible memory-overwrite bug in the display command's insertion of dummy variable names into the expressions it echoes. The offending lines in an example that revealed the bug were display {f in A}: sqrt(cov[f,f]); var x{B}; with length(f) in the "right" range of values for some f in A. In the example, the bug was revealed at the parsing of x in the "var x{B};" declaration. 20030514 Fix a glitch in the "let" command's assigning values to declared suffixes of objectives: with some sequences of commands, an inappropriate error message about "ungenerated ???" appeared. Example: model diet.mod; data diet.dat; suffix foo; let Buy['BEEF'].foo := 23; let Total_Cost.foo := 8.2; #error message 20030517 Fix a bug with "suffix name expr" phrases in variable declarations (introduced 20030204) that resulted in OUT suffix values not being transmitted to solvers. 20030527 Fix a bug in the MS Windows version with option solver settings that start with a driver letter and colon and do not end in .exe. (Work-around: append .exe explicitly.) Fix a bug (introduced 20030204) with "expand _con;" (no subscript), which faulted. Adjust the output of "expand _con;" and "expand _con[n];" so that when complementarity constraints are present, they are only indicated by a ".L" or ".R" suffix on the constraint name. Previously, if, say, the first constraint was a complementarity constraint, then both "expand _con[1];", "expand _con[2];" and "expand _ccon[1];" all produced the same output, except for the decorations ".L" and ".R" in the constraint names shown for _con[1] and _con[2]. Similarly, "expand _con;" showed complementarity constraints twice, distinguished only by ".L" and ".R" decorations. These decorations now appear after rather than before subscripts. 20030602 Fix a rarely visible glitch only known to affect Linux binaries starting with 20030204 (introduced by a compiler update). The example behind this fix gave "invalid refct 0 in opgen". 20030605 Fix memory-overwrite bugs that were possible when an entity with a 0-dimensional slice was instantiated. Scripts involving such declarations were automatically terminated, but instantiations were possible with interactive commands. Reduce the Warning about a 0-dimensional slice to a Caution, so scripts involving such slices may now run. 20030617 Fix an obscure bug with the interaction of problem declarations, loops, certain generic synonym references, and commands that should not depend on the current problem, such as the print command in the following silly example, in which the print command elicited an error message about n not having a value: var x; minimize zot: (x-2)^2; param n; minimize bletch: (x-n)^2; problem foo: x, zot; problem goo: x, bletch; repeat{ print 'Starting repeat loop'; problem foo; solve; let n := _ncons + 2; problem goo; solve; if _ncons > 0 then display _con[1].slack; } until n > 0; 20030625 Fix a rarely seen bug in handling positive values of option presolve_inteps when "solve" or "solution" read a .sol file. If the bug bit, an "option presolve_inteps 0" command appeared, perhaps followed by a complaint about a wrong number of variables. 20030626 Adjust changes made, if necessary, by the "solution" command to $presolve_eps, $presolve_fixeps, and $presolve_inteps to properly read a .sol file corresponding to a .nl file written before changes to those options: multiply the R(x) value shown in the changes of 20000825 by .95 (to put the value roughly halfway between the relevant integer powers of 1.1), then round the result to 3 significant figures. For example, with a suitable solver and problem instance, in write bfoo; solve; option presolve_inteps 37; solution b.sol; the solution command now issues option presolve_inteps 1.04e-06; rather than yesterday's option presolve_inteps 1.0950988861107401e-06; Because of how $presolve_inteps is discretized, "option presolve_inteps 1.04e-6" has the same effect as the default "option presolve_inteps 1e-6". Today's changes also correct a bug in deciding whether to restore $presolve_inteps. On the relatively rare occasions when this bug bit, it might cause reading the .sol file to fail after a surprising "option presolve_inteps..." command was generated and presolve ran again. 20030707 Fix a glitch, introduced 20030204, in handling function evaluations: subscripted variables appearing as arguments were treated as "constant", as were some other kinds of expressions involving variables. For example, in set A; var x{A}; function f; var y = sum{i in A} f(x[i]); minimize zot: y^2; the "var y =" declaration elicited the incorrect diagnostic Caution: y should be "param :=" rather than "var =". 20030724 Fix a bug in handling logical constraints (a forthcoming extension): if $auxfiles or $($solver & '_auxfiles') requested a .row file and the problem contained both logical constraints and objectives, a fault was likely. New feature related to this bug fix: when new option convert_logical_to_algebraic has its default value 1, constraints that appear to be logical constraints because they are surrounded by parentheses but that would be recognized as algebraic constraints without the surrounding parentheses are converted to algebraic constraints during parsing. Specifying option convert_logical_to_algebraic 0; suppresses this conversion. 20031017 Diagnose ": =" instead of ":=" in data tables. 20031112 Fix a glitch (possible segment fault, probably introduced in version 20030204) in "if" expressions involving variables. 20031202 Fix a bug (fault) with the Unix/Linux invocation "ampl 2>file". 20040103 Fix a bug in "write table"'s handling of a possibly subscripted but otherwise unadorned constraint name: if the constraint was eliminated by presolve, its dual value may not have been computed (depending on the preceding commands). Adjust linearization of a nonconvex piecewise-linear term to use max($pl_bigM, 2*max{b in Breakpoints} abs(b), if lb > -Infinity then abs(lb) else 0, if ub < +Infinity then abs(ub) else 0) where $pl_bigM was previously used (with lb and ub the lower and upper bounds on the variable that the piecewise-linear term multiplies). Turning the "8" bit of $pl_linearize on, e.g., by option pl_linearize 9; suppresses this change, at least for now. 20040202 Banish the inappropriate error message "Cannot assign a value to ungenerated ??? p." that was given for set A default {}; var p{A}; let{i in A} p[i].relax := 1; Fix glitch in presolve that caused some initially "nonlinear" variables to remain counted as nonlinear even though simplifications rendered them linear. An example where this happened: set I := 1..3; var x{I} >= 0 <= 10 integer; minimize zot: sum{i in I} i*x[i]; s.t. cb: x[1] >= x[2]*x[3]; fix x[3] := 0; Fix fault in handling models with piecewise-linear terms, all of which get eliminated by presolve, and constraints that would be defining constraints under -S (i.e., option substout 1). Example: var x; var y; var z >= -3 <= 4; s.t. foo: y = (x-1)^2; minimize zot: y + <<1;-1,1>>z; s.t. bletch: z = 2; 20040210: Fix glitch in simplifying nonlinear constraints: a variable multiplied by a variable expression that presolve determined to be zero was still regarded as appearing nonlinearly when at least one other nonlinear expression remained in the constraint. Example: var x{1..3} >= 0; var z >= 0 <= 0; s.t. zot: z*x[1] + x[2]*x[3] >= 3; s.t. zap: x[1] >= x[2] * x[3]; # x[1] was recorded as a nonlinear variable In the example set I := 1..3; var x{I} >= 0 <= 10 integer; minimize zot: sum{i in I} i*x[i]; s.t. cb: x[1] >= x[2]*x[3]; fix x[3] := 0; mentioned in the changes of 20040202, recognize that constraint cb reduces to a bound on x[1]. 20040229 Minor tweak to change of 20040210 to permit deducing stronger bounds in some cases after detecting multiplication by zero has made some nonlinearities go away. Fix a glitch in sequences of the form write ...; shell ...; # obtain a .sol file for the .nl file just written option ...; # change $presolve and perhaps some presolve tolerances solution ...; # read the .sol file If $presolve was increased, the solution command might have inappropriately caused changes to other presolve tolerances. (Such changes are indicated by the echoing of option commands.) 20040422 Fix a bug that might have surfaced after an error message about discarded subscripts immediately followed by "Bailing out after n warnings". Fix a fault in the error message for the circular definition of y in var x; var y{i in 1..3} = if i == 1 then y[3] else x*y[i-1]; minimize zot: y[3]; write 0; Introduce new error message to diagnose scalar "defined" variables that involve themselves, as in "var q = q*sin(q);". As before, such constructs introduce constraints, but now they elicit an error message saying so. 20040515 Fix an obscure bug that gave an equally obscure error message: "unexpected type 0x16e0 in massage()". 20040604 Fix a bug in simplifying nonlinear constraints after presolve fixed nonlinear variables appearing in the constraints. When the bug bit, inequality constraints might have been treated as equalities in some bound computations, possibly resulting in incorrect bounds or even the incorrect fixing of variables, including defined variables. 20040613 Fix a fault possible with "solve" after some (unusual) sequences of commands, such as model foo.mod; data foo.dat; display someconstr; solve; where someconstr is a constraint and the .mod and .dat values are what they claim to be -- files containing a model and data and nothing else. [One often sees the bad practice of including commands in .mod and .dat files.] Note that the [dual variable] value(s) printed for someconstr would all be zero unless someconstr's declaration specified a dual initial guess. 20040821 When the problem size in the student edition of AMPL is exceeded, have the error message say "the student edition of AMPL is limited..." rather than just ""the student edition is limited...". Print NaN values as NaN, regardless of their sign bits. Previously -Nan was sometimes (incorrectly) printed. The sign bit of a NaN is not supposed to have any meaning. Add new system param NaN (which, on systems with IEEE arithmetic, has the value NaN and in general has the value Infinity - Infinity). Fix a bug with discarded subscripts, illustrated by set A; param p{A}; data; set A := a b c; param p := a 1 d 2 c 3 b 4; display p; # correctly complains about p['d'] display p['c']; # incorrectly said "invalid subscript p['d']" 20040827 Fix a bug that might be revealed by complicated recursive set declarations and gave the message "cp_lookup: unexpected hit". Example: set A = {1,2}; set S {s in A} = if s == 1 then {(2,1)} else {i in A, j in A: j <= i+1 && exists{k in A} ((i,k) in S[s-1] || (k,i) in S[s-1])}; display S; 20040902 Fix another bug that might be revealed by still more complicated recursive set declarations and gave the message "cp_lookup: unexpected hit". 20041110 Fix a bug (fault) in handling "{if something} expr" in display commands, e.g., display {if 0 == 0} 1; 20050208 Fix a bug (fault, or worse) in handling default expressions in certain situations with common subscript expressions. An example that gave the surprising error message "set::rehash finds duplicate entry": set A; param p = 27; param q{A} default 1; var x{A}; minimize zot: sum{i in A} q[i]*x[i]; data; set A := a b c d; let {i in A: q[i] > 0} q[i] := q[i]*p; write 0; 20050424 Fix glitch with display statement formatting of 2-D tables in which all row labels are '' (the empty string): the row labels were taken to have width 0 instead of 2 (for the quotes). Example: param n; set I = 1..n; param c {I}; data; param n := 5; param c := 1 16 2 22 3 12 4 8 5 5; option dis*1col 0; option display_transpose -10; # columns did not line up right in the following display command: display{j in {''}, i in I} c[i]; Fix glitch with sets and params constructed from random functions: sometimes the random expressions were surprisingly re-evaluated. Example: set A{1..2} = {Uniform01(), Uniform01()}; display A; display A; # values had changed Fix glitch with {if ...} indexing of objectives and constraints: after an "expand" command elicited a complaint about ... being false and a change that made ... true, an "unexpected ctype" message was possible. Example: var x; param N default 1; minimize foo{if N >= 2}: (x-1)^N; expand foo; let N := 2; expand foo; # gave unexpected ctype 536871429 in compile_objectives Fix a rarely seen bug that manifested itself only under complicated circumstances. 20050503 Fix bug in reflecting bounds on defined: under conditions hard to predict (but apparently not often seen), some constraints were inappropriately relaxed. To see if the bug bit, print min{i in 1.._ncons} _con[i].slack; which should be non-negative but for roundoff. Work-around: option substout 4; 20050530 Fix a bug with the read command's handling of iterated arguments: a fault (or worse) was possible under conditions hard to predict. An example where the bug caused a fault: set A dimen 2; param E{A}; set I = setof{(i,j) in A} i; set J = setof{(i,j) in A} j; param y{I}; let A := {1..699,1..3}; read{i in I} y[i] < 'y'; for{j in J} read{i in I} E[i,j] < ('pred' & j); #fault If a "read" command issued in interactive mode encounters end-of-file, permit further interactive input. Fix bug ("level has impossible levbits") with iterated print, printf, and read commands with redirections that vary with the iteration. Example: print{i in 2..3}: {j in 1..3} j^i > ('x' & i); Fix a bug possible only in interactive mode after an error message (such as "no value for ...") interrupts processing. Example: set A dimen 2; param E{A}; set I = setof {(i,j) in A} i; set J = setof {(i,j) in A} j; var x{J} >= 0; set Ap = {(i,j) in A: E[i,j] < 0}; minimize ssq: sum{i in I} (sum{(i,j) in Ap} E[i,j]*x[j])^2; s.t. convex: sum{i in J} x[i] = 1; let A := {1..208,1..5}; solve; #ampl: Error executing "solve" command: #error processing set Ap: # no value for E[1,1] read{j in J}: {i in I} E[i,j] <('foo' & j); solve; #Segmentation fault 20050614 Fix bug with "redeclare": things were not always recomputed after "redeclare" should have changed their values. Example: set I := 1..3; set J := 1..4; set B within {I,J}; param Maj = floor(card(J)) + 1; set BV = {i in I: card{(i,j) in B} >= Maj}; data; set B := (1,2) (1,4) (2,2) (2,3) (2,4) (3,1) (3,3) (3,4); display Maj, BV; redeclare param Maj = floor(card(J)/2) + 1; printf "\nUpdated Maj = %d\n", Maj; display BV; # said "set BV := ;# empty" # rather than "set BV := 2 3;" Fix bug with a slice on a set not yet given a value. Example: set A dimen 2; param E{A}; set I = setof {(i,j) in A} i; set J = setof {(i,j) in A} j; var x{J} >= 0; set Ap = {(i,j) in A: E[i,j] < 0}; minimize e: sum{i in I} (sum{(i,j) in Ap} E[i,j]*x[j])^2; data; set A := a b a c a d b c c d; expand e; # no value for E['a','b'] data; param E := a b 1.2 a c -2.3 a d 3.4 b c -4.5 c d 5.6; expand e; # was "0"; should be (-2.3*x['c'])^2 + (-4.5*x['c'])^2; 20050624 Fix long-standing bug in the "display" command's handling of entities indexed over an ordered set of arity 1: the ordering was ignored. Example: set WEEKS ordered; param avail{WEEKS} >=0; data; set WEEKS := 27sep 04oct 11oct 18oct ; param avail := 27sep 40 04oct 40 11oct 32 18oct 40 ; display avail; # listed 04Oct first rather than 27sep Fix glitches with complementarity constraints. Nonlinear defined variables were sometimes mishandled (a bug that crept into version 20040422). Assignments of dual values to complementarity constraint names sometimes led to subsequent faults (or worse). Note that when ccname is an unsubscripted complementarity constraint, "let ccname := value;" is the same as "let ccname.Ldual := value;". 20050702 Fix a bug with recomputing reduced costs after a change to $abs_boundtol or $rel_boundtol that could affect the reduced costs: dual values for constraints eliminated by presolve were sometimes miscomputed. The bug was revealed by a command sequence of the form option abs_boundtol 1e-5; solve; display _con; option abs_boundtol 0; display _con; option abs_boundtol 1e-5; display _con; in which "display _con" caused reduced costs and dual values for constraints removed by presolve to be computed. Note that to get correct dual values for constraints removed by presolve when using an interior-point solver, it's necessary to set $abs_boundtol or $rel_boundtol to a suitable positive value. These options are not described in the AMPL book, but are described in the entry for 19931005 of /netlib/ampl/changes. Note also that not all changes to $abs_boundtol and $rel_boundtol cause recomputations; AMPL records ranges of values for these options that have the same effect and avoids recomputations for changes that do not matter. 20050711 Fix a bug with assigning values to variables by "read table" or by calls on imported functions with OUT arguments: absent a previous "solve", the newly assigned values were not passed to solvers. 20050818 Fix a bug (e.g., fault) with iterated printf commands whose format varies with the iterate. Example: printf{i in 1..10} (if i mod 2 == 0 then "%s\n" else "%s "), i; # faulted 20050904 Fix glitch with use of "integer" or "logical" in recursive definitions. Example: param factorial{i in Integers} integer = if i <= 0 then 1 else i*factorial[i-1]; display{i in 1..4} factorial[i]; gave the surprising error message "unexpected type 0x6500 in cpexpr()". 20050908 Fix a botch (fault) introduced 20050818 with simple iterated printf commands, such as printf {i in 1..2} "%d\n", i; 20050910 Adjust error messages for recursive definitions to show what basic entity was being instantiated (the one that directly or indirectly referenced the recursive set or param). Misleading error messages were previously possible (as illustrated by a complicated example). 20050927 Fix a bug in handling nested "if" statements without {} enclosing "then" parts and no outer "else" part: following statement was only executed if the outer "if" was true. Example: if 1 == 2 then if 3 == 4 then display 5; display 6; # was not executed Fix a bug with "reset data": supplying larger sets in a data section sometimes led to a fault in subsequent commands. Fix a glitch in simplifying nonlinear expressions that end up being linear: simplifications from expr1 * expr2 to constant * linear were sometimes treated as nonlinear, possibly resulting in the solver linear constraints marked "nonlinear" or perhaps in presolve not making as strong deductions as it should. 20051010 Fix a glitch in "show x;" where x is a nonlinear defined variable: a double minus sign was sometimes unnecessarily produced. Example: var x; var y = x^2 + 1; show y; # said "var y = --(x^2 + 1)" instead of just "x^2 + 1". 20051012 Fix a bug in handling conditionally evaluated expressions: when such an expression occurred more than once, it was sometimes evaluated even though a guard condition implied it should not be evaluated. Example: set A; set B ordered; set C; param p{A,C}; param q{A}; param r{i in A} = sum{j in B: j in C && q[i]/p[i,j] < 2.4} ord(j)*(q[i]/p[i,j]); data; set A := a b c; set B := x y z; set C := x z; param p: x z := a 1.3 2.4 b 4.9 0.3 c .23 .45 ; param q := a 5.2 b 6.7 c .23; display r; # gave "invalid subscript p['a','y']" 20051016 Fix some bugs with logical constraints (a forthcoming feature): "a ==> b else c" was not handled in presolve, and declaring a suffix after declaring a logical constraint led to a fault. 20051110 Fix a bug with problems having both logical constraints (a forthcoming feature) and defined variables: incorrect .nl files were written. 20051119 Fix a bug with suffix declarations or some redeclare statements appearing after function declarations: a fault or other confusion was possible. 20051128 Fix glitch in writing .row files: if the model contained both logical constraints and defined variables, wrong name were emitted for the logical constraints. Fix bug in non-iterated read commands that start with an iterated item, followed by another item, such as read {i in A} p[i], q; The second and subsequent items (q in the example) were ignored. 20051202 Fix bug in handling suffixes on constraints when logical constraints are present. 20060314 Fix a bug with handling end-of-file in a read command, i.e., in "read ... = 4GB (a mostly invisible change; for solvers, this requires ASLdate >= 20060122). 20060331 Fix a typo in an error message ("errror" rather than "error"). Fix a bug (possible fault, or worse) with "expand" on 64-bit systems. 20060408 Fix a bug handling long comments (more than 8192 bytes long) that start with #. Fix bugs handling "let _var[...].relax := ...;". Sometimes the assignment appeared not to take effect; other times, iterated such let commands ran surprisingly slowly. Fix bugs in "show;" introduced 20060314. 20060430 Arrange for presolve to turn logical constraints into algebraic constraints (when deductions make this possible), and then to process the new algebraic constraints. New builtin params: _nlog_algcons = number of logical constraints turned by presolve into algebraic constraints _npre_log_algcons = number of such constraints subsequently eliminated by presolve. These numbers also appear in the output for "option show_stats 3;". 20060506 Fix a glitch ("unexpected subtype 13 in prset") with "show" of a set with an "ordered by" clause. 20060524 Fix a bug in handling some complicated expressions where an inner indexing set depends on outer indexes and the inner indexing set applies to an expression that only involves dummies over the inner indexing set. Examples satisfying this description are not hard to construct, but finding a simple such example where the bug bit is not easy. Here is such an example. (Writing the example more sensibly avoids the bug.) set I = 1..3; set A = {I,I}; set B = {I,I,I}; set C{B} default A; param p{B} default 0; param q{A} default 0; let{(i,j,k) in B: max( # faulted {(a,b) in A: a == i and b == k} q[a,b], {(a,b) in C[i,j,k]} q[a,b] ) > 0} p[i,j,k] := 1; display p; #empty 20060626 Fix glitches in "option show_stats 1" introduced 20060430: the number of linear variables was sometimes not reported or reported incorrectly, and "0 constraints" was omitted when there were no constraints. Fix longstanding bug in "option show_stats 1": for nonlinear constraints, "linear nonzeros" appeared instead of just "nonzeros". 20060725 Have "display" honor $display_precision when printing symbolic parameters having numeric values. Fix bug (probably introduced 20060314) in the "show" command's labeling of environments, suffixes, and tables. When "expand" or "solexpand" prints out a constant objective, show its value as "number" or "-number" rather than "0 + number" or "0 - number". 20060905 When x is a variable, have "reset data x;" discard x's current value (and, as before, permit new default data for x's initial value in a data section). If x has a random default expression, it gets resampled when x is next needed. Previously, "reset data x" worked this way until x was assigned a value by "let" or "solve", after which x retained the assigned value. Fix some glitches in _slogconname when defined variables are present. Fix possible fault after diagnosing "0-dimensional slice". Omit warning, "This recursive set declaration must specify dimen before := or default." and instead infer the appropriate dimen. Fix bug (incorrect diagnosis of implicit definition) in checking membership in some complicated recursively-defined sets. Example: set A; set P within {A,A}; set S{n in 1..card(A)} dimen 2 = if n == 1 then P else S[n-1] union setof {k in A, (i,k) in S[n-1], (k,j) in S[n-1]} (i,j); data; set A := a b c d e f; set P := a c b c c d e f; print if ('a','c') in S[3] then 'yes' else 'no'; # said "sorry, no implicit definitions: cannot evaluate S['a']" 20060912 New option csvdisplay_restrict: with $csvdisplay_restrict at its default value, an undocumented restriction on csvdisplay is henceforth relaxed. Specifying "option csvdisplay_restrict 1;" enforces the restriction in question, namely that csvdisplay issues an error message (meant to help debug GUIs that use csvdisplay) when asked to display more than one table. Example: set A{i in 1..4} = i .. i^2; csvdisplay A; # used to be disallowed option csvdisplay_restrict 1; csvdisplay A; # gives the old behavior: # complains "csvdisplay would emit 4 tables." 20060915 Fix glitch with "reset data x;" where x is a variable or constraint with random initial value: in the sequence solve; reset data x; solve; the initial value of x was not resampled for the second solve. Other commands that required x's value did (and still do) cause resampling. Fix bug (fault or surprising error message) with "printf "%f", 1e75;". 20060930 Fix a bug (fault or worse) in cleaning up after an error message when regenerating an indexed variable, constraint, or objective after some change made regenerating necessary. 20061005 Fix glitches with imported functions: 1. If-then-else expressions in constraints and objectives with imported functions appearing only in the "then" or "else" expressions could result in a complaint about unavailability of the functions. They might be available to the solver, so no complaint should appear. Example: function exact; param p symbolic; var x >= 0 <= 10; minimize zot: (if num(p) == p then p else exact(p))*x; data; param p := '1234567890123456789'; write 0; # "Can't invoke unavailable function exact." 2. If an if-then-else expression in an argument to an imported functions evaluated by the solver had a non-constant numeric "then" expression and a string-valued "else" expression or vice versa, confusion resulted in that a numeric expression might appear where a symbolic one was expected. Example: function goo; var x; minimize zot: x*goo(x, if x >= 3 then 'abc' else x+10); # new diagnostic: "solve" or "write" now says # Cannot convert a numerical "then" or "else" to string # in if-then-else expressions that solvers must evaluate. 20061007 Fix bug in handling chains of comparisons involving more than 2 comparisons. Example: print{i in 1..2}: if 1 < i < i^2 > 3 then 'yes' else 'no'; # printed "no", then faulted 20061023 Fix an obscure bug that gave rise to error messages such as presolve has k = 167168, P.nfc = 166781 Tentatively introduce option presolve_logfile with possible values "filename" to write file filename anew each time presolve runs; ">>filename" to append to filename each time presolve runs; and "-" to write to the standard output file (stdout). When $presolve_logfile is not "" (its default value), each time presolve updates a bound, the constraint and variable involved and updated bound are written to $presolve_logfile. New builtin symbolic param _table_errmsg records the last error message from the most recent "write table" or "read table" command, and is set to "" if the last such command had no error. New option table_errbreak determines whether "write table" and/or "read table" commands report errors and terminate processing of commands, or suppress error reports and simply record the last error in _table_errmsg. $table_errorbreak (default 0) is the sum of 1 {suppress error reporting for "write table"} 2 {suppress error reporting for "read table"} 4 {for iterated "read table" and "write table" commands, do all iterations despite errors} Hitherto, $table_errbreak = 4 was the old behavior; now the default value 0 stops iterated table commands at the first error. 20061030 Fix a glitch in the display command's use of $n to stand for the heading of column n: sometimes the heading was '$n' rather than simply $n (unquoted), but the quotes were not accounted for, causing the headings not to line up with the columns beneath them. Now the erroneous quotes are gone. Example: set A; set B; param p{A,B}; data; set A := 'pumpkin pie' 'ice cream' 'chocolate sauce' 'vanilla cream' 'apple pie' strawberries; set B := 'New York' 'San Francisco' 'San Diego' 'Washington, D.C.'; param p : 'New York' 'San Francisco' 'San Diego' 'Washington, D.C.' := 'pumpkin pie' 12 13 14 15 'ice cream' 22 23 24 25 'chocolate sauce' 32 33 34 35 'vanilla cream' 42 43 44 45 'apple pie' 52 53 54 55 strawberries 62 63 64 65; option display_width 40; display p; # heading ended with "'$1' '$2' '$3' '$4' :=" # instead of with "$1 $2 $3 $4 :=" 20061102 Fix another bug with recursive set definitions. A fault revealed the bug, but wrong results without a fault might sometimes have been possible. 20061121 Add messages distinguishing "Solution determined by presolve" from "All variables fixed" and "All relevant variables fixed". (In the latter case, some unused variables are not fixed.) All these cases still get solve_result_num = 99. Have main() call exit rather than returning, to prevent a fault on an odd version of Linux. 20061130 Fix glitch with suffixes on random variables (a forthcoming addition). 20070222 Fix bug apparently introduced 20061005 in calling user-defined functions: numeric dummy arguments, such as i in "f({i in 1..3} i)", were passed to solvers as strings rather than numbers. 20070301 Fix glitch in "read" command introduced 20060314 that could cause an immediate "end of file" error message. Tweak to $csvdisplay_restrict: when $csvdisplay_restrict mod 2 == 1, provide a more detailed error message when csvdisplay would emit more than one table, unless $csvdisplay_restrict mod 4 == 3, in which case the former, one-line error message appears. 20070312 Fix bugs in handling variables declared "in Set_expression": integer and binary were not always properly distinguished when such variables could be turned into integer or binary variables; with "option presolve 0", the "in Set_expression" was ignored; and at one point in presolve, a test for reducing the domain of a variable using the "in Set_expression" was applied too soon, possibly causing variables to be fixed inappropriately. Examples giving rise to these troubles had the form var x in {0, 1, 2}; # or in {0, 1} for the third bug var y >= 1, <= 3; maximize obj: y; subject to r: y <= 2 * x; 20070317 Fix a bug with option single_step: if an error (such as a bad subscript) was reported during single-stepping, things were sometimes not properly cleaned up, possibly resulting in a subsequent fault. Addition to error messages for commands: the current _cmdno (the count of commands executed) is now reported unless input is from the standard input. For the standard input, setting option stdin_offset to 1 causes the input offset and _cmdno to be reported in error messages for commands. New option single_step_cmdno causes $single_step to be set large enough to cause single stepping when _cmdno == $single_step_cmdno. This provides a way to repeat the execution (after "reset;" or a fresh invocation) and enter single-step mode just before an error, to permit looking at current values, which might shed light on the error. 20070410 Add " 7 rand random variable in current problem" to $astatus_table. Fix bug in variable.astatus when there are unused defined variables. Example: var x; var y; var u = 4*x + x^2; # unused var z = 3*x^2 + y^2; var w; minimize zot: z + w; s.t. bletch: exp(z) <= 4.2; display _varname, _var.astatus; # showed "sub" rather than "unused" for u 20070505 Fix bug during "reset;" with handling at_exit() registrations by imported functions: they were lost. Fix bug in handling "table" declarations appearing before a suffix declaration (including implicit suffix declarations that result from "solve" commands in which the solver returns a new suffix), or that appear before certain "delete" commands: confusion, such as a fault or surprising error message, could result. Example: set I = 1..3; var p{I}; table pt OUT 'pt.tab': {i in I} --> [eye] , p[i]; suffix zork; write table pt; # "unexpected type..." Fix bug with multiple suffix declarations appearing after some other declarations and a delete command: a fault or other confusion was possible. 20070518 Fix a bug in presolve's handling of variables declared "in Set_expression". Such set expressions in general are unions of intervals (with point values included as degenerate intervals). If presolve deduced a bound that would eliminate more than one interval, it botched a bound update. Example: var x in {0, 2, 4, 8}; s.t. bletch: x <= 2; # the bug changed 2 to 4 maximize obj: x; 20070522 Fix handling of string expressions in "commands (...);" commands so ... can involve params declared with "= value" but not yet used in a command. 20070612 Tweak printing of sets under complicated conditions, such as error messages involving next() and prev() under obscure conditions in which the sets are now shown as a list of elements, enclosed in braces. Previously the braces were omitted and the elements printed one per line, which was confusing. 20070618 Fix the bug noted in the second "solve" comment of var x; minimize f: func(x); let x := ...; solve; # with no primal or dual values returned, e.g., gjh redeclare minimize f: another_func(x); solve; # value assigned by "let" to x not transmitted. 20070712 Fix bug in handling imported functions that specify a domain for INOUT arguments with "..." indexing. For example, function f(Reals IN, ...(Reals INOUT)); encountered the bug when invoked, but function f(Reals IN, ...(INOUT)); did not. Fix a bug in simplifying certain complex logical expressions that led to the error message "unexpected type 0x11e03 in lsimplify()". Fix a bug with named problems that led to the error message "named_genall called". Fix a bug with "delete" and "purge": when applied to variables, constraints and objectives, they did not update problem declarations, which could have led to a fault or other confusion. 20070802 Fix a bug (infinite loop) in "redeclare set ..." when neither the old or new set is subscripted. Example: set S; var x{S} >= 0 <= 1; set F = {i in S: x[i] in (0,1)}; #error message redeclare set F; # looped Fix a glitch with "show p;" for a param p that had data supplied in a data section. Example: param p; data; param p := 1.7; show p; # mentioned "random" Fix a performance bug: reduce overhead when params and sets without constraints on their values are updated. 20070825 Fix a fault in handling "ordered by ..." in certain situations where ... is a set expression, such as a subscripted set. 20070903 In the breakpoint and slope lists of piecewise-linear expressions, i.e., the ... of <<...; ...>>, permit iterating parenthesized lists of expressions, as in function calls. Here is an example that is now acceptable, but formerly elicited a syntax error message: var x in [-15,15]; minimize zot: <<{i in 1..3} i^2; {i in 1..2} (-1,1)>> x; 20070923 Permit Infinity (as well as NaN) as values in data sections. Disallow deleting some suffixes not previously noted as system suffixes. System suffix stage was quietly introduced a while ago, for use with random variable declarations (a forthcoming extension). Permit explicit declaration of suffix stage if no random variables have been declared, which disallows subsequent declarations of random variables and permits using stage as an ordinary declared suffix. 20071007 Adjust computations behind "Too much memory used" message to make the message more accurate when sbrk(0) proves to be unreliable. 20071111 Fix bug in handling variables unused in the current problem when defined variables (also unused) were present that depended on the unused variables, the unused variables and their defined variables were mishandled, such that commands involving them might see incorrect values. Example: var x; var g = x^3 - x^2 - x - 1; var gp = 3*x^2 - 2*x - 1; let x := 2; display x, g, gp; # OK so far let x := x - g/gp; display x,g; # botched values (It is more appropriate to use "param" rather than "var" in this example; doing so would have avoided the bug.) Fix a bug with redeclarations: a fault was sometimes possible. Example: set I; param p{I}; var f; param q; minimize zot: sum{i in I} (f*q^i - p[i])^2; s.t. zotc{i in I}: p[i] == f*q^i; delete zot; set J; redeclare s.t. zotc{i in J}: p[i] == f*q^i; # fault 20071121 Fix bug with "ordered by": a fault was possible under the "right" conditions, as in this example: set A ordered; param p{A}; set S ordered by A default{i in A: p[i] == 0}; data; param :A: p := a 0 b 1 c 0 d 2; display S; # faulted 20071222 Banish an obscure error message "unexpected subtype 23 in sforce". Fix fault with "let Set := (expression involving the Set)", e.g., set A default {}; for{i in 2..4} { let A := A union {j in i .. i^2: j in i+2..i+3 diff A}; display A; } 20080102 When a constraint implies a bound on a variable and integrality of the variable causes a further adjustment to the bound, do not treat the constraint as tight when inferring dual values. Before this change, constraints eliminated by presolve were sometimes given nonzero dual values, even though they were slack. Tweak ODBC table handler so (e.g.) with MS SQL Server it will choose a better type for numeric data. 20080128 Fix a bug giving an erroneous "no data" message in the following: set I = 1..3; minimize zot; var x{i in I} integer >= 0 obj zot i; set S; s.t. zap: sum{i in S} x[i] = 1; let {i in 2..3} x[1].relax := 1; # no data for S let S := 1..2; solve; 20080202 Fix another bug giving an erroneous "no data" message under some circumstances when a defined variable is involved in column generation. Example: set I = 1..3; minimize zot; var x{i in I} integer >= 0 obj zot i; var y obj zot 3 = x[3] + 1; set S; s.t. zap: sum{i in S} x[i] = 1; let {i in 2..3} x[1].relax := 1; # no data for S let S := 1..2; solve; Fix a fault arising in a complicated use of defined variables. 20080207 Fix a bug that crept in sometime in printing certain expressions, such as iterated piecewise-linear terms. The bug gave rise to error message "unexpected type 0x1e600 in prex()". Example: var x{i in 1..3} := i; display {i in 1..3} <<{j in 1..i} .5*j; {j in 0..i} (-j)^j>> x[i]; 20080216 Fix a bug (fault or worse) connected with a subscripted defined variable whose indexing set changes size sufficiently to elicit the bug. This also corrects a related (so far not seen) bug with "expand" and ".astatus" of partially dropped constraints. 20080219 Fix a bug in handling iterated slopes and breakpoints in piecewise-linear terms. When it bit, the bug led to a surprising error message, or worse. 20080306 In interactive mode with stdin on the current include stack, have "include 'unavailable'" return to the command prompt (to reading stdin) when file 'unavailable' does not exist. Fix a bug (fault) in printing the error message "bad subscript (use before def?) for defined var ..." under the "right" conditions. 20030307 New variant -Rw of command-line option -R (described in change-log entry for 19970814) disables output redirections ("> filename" and ">> filename"). 20080312 Fix bug in expressions involving the dual value of a subscripted constraint after earlier subscripts of the constraint were dropped, and in writing initial guesses for such constraints in .nl files. New option ampl_libpath specifies a sequence of directories, one per line (analogous to $ampl_include) in which to look for libraries mentioned in "load" commands. When $ampl_libpath is empty or contains only white space, the "load" command looks in the current directory. (The "unload" command looks for names mentioned in previous "load" commands and quietly ignores names not found; "reload" is a combination of "unload" and "load", so it honors $ampl_libpath when doing its "load". Note that $AMPLFUNC and the builtin sets _AVAILFUNCS, _AVAILFUNC2, _LIBS are updated by "load", "unload", and hence "reload".) 20080314 Fix another bug in writing dual initial guesses to .nl files when there are subscripted constraints, some of which have been partially dropped. An invalid .nl file could result. 20080423 New builtin params _last_solno, _primal_solno, _dual_solno, _status_solno: _last_solno is incremented at each "solve" and "solution" command that reads at least some solution information. If any primal or dual variable or solver status (.sstatus) values are read from the .sol file, _primal_solno, _dual_solno, and/or _status_solno, respectively, is/are set to the new _last_solno value. If "solve" results in "Solution determined by presolve" or "Infeasible constraints determined by presolve", these parameters are also updated. 20080616 Adjust reporting of memory used to avoid confusion under some 64-bit versions of Linux where there can be large gaps in the addresses returned by malloc(). This change affects the memory usage reported with "option times 1;" and "option gentimes 1;". Fix a bug in handling defined variables fixed by presolve: linear variables were sometimes reported as nonlinear. Example: var x; var y = x + 1; var z = 2*x + y; s.t. c: z = 12; var w; minimize zot: w; # erroneously reported to have # a nonlinear variable and nonlinear constraint 20080629 Fix a bug in use of variable values in set expressions (which can only happen in commands): if previous changes caused indexing of some variables to change and nothing had since caused the "collect" phase to run again, wrong values were sometimes used. In the very complicated example that led to finding this bug, an inserted "display" changed some behavior (because it caused the "collect" phase to run). With the fix and "option times 1" to show phases, you will still not see "collect" running before a command that does not need it, but correct values are now computed. Further adjustment to reporting of memory used to avoid negative values. Write "Solution determined by presolve" messages to stdout rather than stderr. New builtin solve_result_num values: 999 ==> error running solver (nonzero solve_exitcode); 998 ==> nonexistent .sol file; 997 ==> error in .sol file; 996 ==> error in restoring state to read .sol file. Treat any of these as exit code 1024 from the solver as far as option solve_exitcode_max is concerned. (The default value of 0 for $solve_exitcode_max will then cause a break in the execution of commands.) This applies to both "solve" and "solution" commands. (Of course, solve_result_num cannot be set to 999 by a "solution" command.) 20080701 In some cases of failed "solve" and "shell" commands, report where in what input file the error occurs. 20080703 Tweaks for stochastic programming extensions (to be described later). 20080717 Fix a bug introduced in version 20080616 in the handling of nonlinear defined variables, e.g., those used in both constraints and objectives. Example: set I = 1..2; var x{I} := 1; var y{i in I} = x[i]^2 + i; minimize zot: sum{i in I} (y[i] - 4)^2; s.t. foo: sum{i in I} y[i] <= 7; 20080804 Fix a bug introduced in version 20060905 in which "update data", like "reset data", discarded current values. Now current values are gain retained by "update data". Example: set I = 1..2; var x{I}; param p{I}; minimize zot: sum{i in I} (x[i] - p[i])^2; data; param p := 1 1 2 2; solve; update data; data; param p := 2 5; print sum{i in I} (x[i] + p[i]); # printed 6 rather than 9 New params (mainly for use in debugging carelessly written models): _backslash_map default 0; # 0 ==> no mapping # 1 ==> map \\ to / _filename_case default 0; # 0 == retain given case # 1 == force to lower case set _file_prefixmap dimen 2; # for (a,b) in _file_prefixmap, map prefix a to b Note that under MS Windows, you have always been able to use / instead of \ in file names. Use of \ causes gratuitous unportability. 20080821 Fix a glitch that caused mention of unused imported functions in .nl files when expression simplification (such as multiplication by zero) caused removal of all references to the functions. Example: function ginv; function hypot; set I = 1..2; var x{I}; var y{I} >= 0; s.t. bletch: sum{i in I} y[i] <= 0; minimize foo: hypot({i in I} (x[i] - i)) + y[1]*ginv(y[2]); # ginv was listed as required in the .nl file, # even though presolve deduces y[1] == 0 and no actual # reference to ginv appears in the .nl file. Fix a glitch with "delete option ...". Example: option zork 'some value', zork; delete option zork; option z*; # said option zork ''; # not defined # after the bug fix, "option z*;" does not mention zork at all. 20080907 Arrange for printf "%+011.0f\n", -12345; to print "-0000012345" rather than " -12345". Extend printf to accept the C99 formats %a and %A. 20080925 Fix two obscure bugs that should be evident if encountered. (One only appeared on some PowerPC systems; the other was elicited in testing a forthcoming extension, not yet announced.) 20080926 Add option shell_exitignore to suppress the message "exit code n" from "shell" commands when n <= $shell_exitignore. (On Unix and Linux systems, n = shell_exitcode/256.) 20081026 Fix a bug in handling two-sided constraints involving defined variables. If the constraint implied lower and upper bounds on a defined variable and something else implied a stronger lower bound on the variable, the constraint was mistakenly dropped during presolve, even though the side that implied the upper bound should have been retained. Add "q" as synonym for "quit". 20081120 Fix a bug with recording drop/restore state when a problem is declared, the drop-restore state then changed, and a second problem is declared. The modified drop-restore state was not saved with the first problem. Example: set S; var x{S} >= 0; problem P1: x; fix x[2]; problem P2: x; # the fixing of x[2] in P1 was lost A work-around is to declare problems before explicitly modifying the drop-restore state. 20081208 New option mpsfile_numwidth (default 12) controls the field width and format of floating-point values written by the -om output option (for MPS files, which are to be discouraged, as they are slow to read and write and omit important problem details, such as the sense of optimization). Positive values specify that field width; negative values specify fieldwidth -$mpsfile_numwidth with nonnegative values having an initial space, so both x and -x are printed to the same number of significant digits. Specifying "option mpsfile_numwidth 0" causes numbers to be written to full precision, the same as printf format %.g (in AMPL -- a sensible departure from the unfortunate C99 provision that %.0g be treated as %.1g). Some solvers may not cope properly with values of abs($mpsfile_numwidth) > 12. New option no_hexfp. When $no_hexfp is 0 (its default), C99-style hexadecimal floating point constants are recognized: they are strings that start with 0x or 0X followed by a string of hexadecimal digits possibly containing one decimal point, optionally followed by an exponent part consisting of p or P, an optional sign (+ or -) and a nonempty string of decimal digits (the power of 2 by which to multiply the hexadecimal value preceding the exponent part). When $no_hexfp is 1, hexadecimal floating-point values are not recognized, thus restoring AMPL's previous behavior, in which hexadecimal floating values without a decimal point could be used as normal identifiers, and general hexadecimal floating-point values could appear unquoted in data sections and be treated as symbolic values. One reason for interest in hexadecimal floating-point numbers is that they are faster to read and write than decimal values. Note that the printf formats %a or %A can be used to write hexadecimal floating-point values. We expect most AMPL users to be unconcerned with hexadecimal floating-point notation, but on occasion a few may find it convenient to be able to read and write them. New builtin set _LOCAL_OPTIONS is set initially to most of the option values that AMPL provides, except for ones found in the incoming environment, and except for AMPLFUNC amplfunc0 objective_precision version which are always exported, as some solvers use them. When carrying out a "solve" or "shell" command, AMPL does not pass to the invoked program environment variables whose names appear in _LOCAL_OPTIONS. This set is retained across reset commands and can by modified by "let" commands. To restore the old behavior of exporting all options, invoke let _LOCAL_OPTIONS := {}; initially. 20081210 Fix a bug with _obj[...]: wrong values were reported. New command _solexpand, a variant of solexpand meant for program rather than human consumption, and only useful (so far) for linear constraints and objectives. For each constraint or objective treated, it writes a one-line header, followed by linear terms, one per line. The header has the form What sno bt nlin isnl bound(s) name where What is one of Min for an objective to be minimized, Max for an objective to be maximized, or Con for a constraint; sno is the objective or constraint number as seen by the solver: 0 for the first, 1 for the second, etc.; bt is 0 for objectives and for constraints is (if the constraint's lower bound is finite then 1 else 0) + (if the constraint's upper bound is finite then 2 else 0); nlin is the number of linear terms that follow; isnl is 1 if the objective or constraint also has a nonlinear part and is 0 otherwise; if 0 <= bt <= 2, bound(s) is a single number, either the objective's constant term or the constraint's one finite bound; if bt is 3, then bound(s) is two numbers: the constraint's lower bound followed by the constraint's upper bound; finally name is the possibly subscripted name of the objective or constraint in question. The nlin lines for linear terms that follow the header line have the form varno coef where varno is 0 for the first variable that the solver sees, 1 for the second, etc., and coef is the coefficient of the variable. Floating-point numbers by default are written by format %a (i.e., C99-style hexadecimal floating-point format, which now can be read by the strtod routine in the AMPL/solver interface library); if $no_hexfp is set to 1, floating-point numbers are written with format %.g (i.e., full precision decimal). 20081213 Fix a bug in checking for bad subscripts that led, under complicated conditions, to the message "sub_check bug!". Example: param N integer; set A := 1 .. N; param p{A} >= 0; data; param N := 4; param p := 1 2.5 2 3.4 3 4.7 4 8.9; display N, p; let p[3] := p[1] + p[2]; # with p[3] changed to p[2] on the line above # failed to report bad subscripts below let N := 2; display N, A; let p[1] := p[1] + p[2]; # sub_check bug! display p; 20090101 Fix a bug (fault) in the error message for an invalid subscript in a "while" or "until" clause in a "repeat" command. Example: param p{i in 1..2} = i; param q default 1; param c default 0; repeat until p[q] > 3 { let q := q + 1 } # faulted In error messages, report _cmdno to full precision even when it exceeds 1e6. 20090130 Fix bug (fault) in handling empty sequence of commands enclosed in braces in a file read by "commands" command within a loop. 20090209 Complain when a var declaration specifies the variable to be in a disjoint union of intervals involving an infinity. 20090316 Fix a couple of bugs with "break" commands in a file read by a "commands" command and pertaining to a loop in which the "commands" command appears. Fix a glitch with "write" commands involving a string expression in parentheses: the entities in the expression were not brought up to date. Example: param q symbolic = 'abc'; var x; minimize zot: x^2; write ('g' & q); # complained about q having no value Fix a bug in processing "param-data-alternate" tables (see p. 476 of the AMPL book) tables with a subscripted set-name. An erroneous error message about the subscript was given. Extension: in data sections, after a : ... := header, permit ": :=" (which may be written "::=") to reuse the same header. Example: set S dimen 3; param p{S}; data; set S := (a,x,1) (a,x,2) (a,y,1) (a,y,2) (b,x,1) (b,x,2) (b,y,1) (b,y,2); param p [a,*,*] : 1 2 := x 1.1 2.2 y 3.3 4.4 [b,*,*]::= x 5.5 6.6 y 7.7 8.8; display p; Extension: in data sections, permit a "param-data-alternate" header that specifies a set-name and one parameter name to accept ": ... :=" tables instead of a ":= value-item ..." list. Example: set S dimen 3; param p{S}; data; param :S:p [a,*,*] : 1 2 := x 1.1 2.2 y 3.3 4.4 [b,*,*]::= x 5.5 6.6 y 7.7 8.8 [c,*,*]::= z 9.9 0.2; display p, S; 20090323 Fix a glitch that gave rise to error message "unexpected nonvariable type ... in eput". Example: var x >= 0 <= 0; var y; minimize zot: sqrt(x*(x-y)+y); solve; # unexpected nonvariable type 8194 = 0x2002 in eput New error message "The current problem is empty." appears in response to a "solve" or "write" command after a new, empty problem declaration not followed by any subsequent variable, constraint, or objective declarations. Previously one of the messages "All constraints and objectives dropped." or (if some defined variables had been declared) "Solution determined by presolve." appeared in this context. Examples: model diet.mod; data diet.dat; problem zork; solve; # said "All constraints and objectives dropped." and a defined-variable variant model diet.mod; data diet.dat; var Buysum = sum{j in FOOD} Buy[j]; problem zork; solve; # said "Solution determined by presolve." 20090327 Print a warning and ignore "reset data" and "update data" commands applied to builtin params and sets. This fixes a fault with, e.g., reset data _solve_time; solve; #faulted 20090508 This is a note about a longstanding feature that apparently we missed documenting: with interactive input, "/*" leads to prompt "*ampl:" until "*/" appears in the input. In general, the prompt the prompt that would otherwise appear, preceded by *. Fix a bug (fault) with min({i in S} expr(i)) and max({i in S} expr(i)) in contexts where the expression could be discarded after use, as in the following example: set S; param p{S}; data; param :S: p := a 1 b 2.5 c 2.4; display max({s in S} p[s]); Workaround for the above example: display max{s in S} p[s]; Fix a bug with certain references (e.g., to .sno in the example below) whose values were needed before something else forced full instantiation of the problem. Example: var x; minimize zot{i in 1..2}: i*x^2; display zot.sno; # faulted 20090611 Fix bugs (likely causing faults) with partially dropped complementarity constraints appearing in problem declarations and expand commands. 20090726 Add option allow_NaN (default 0). If $allow_NaN is positive, objective and constraint values are set to NaN in the face of a numeric evaluation error that depend on current variable values, rather than giving an error message. Fix a bug that caused error message "tva top error" when it bit. 20090811 Fix a fault with a complicated expression mistakenly using & (the string concatenation operator) rather than && (logical and). Add a warning about comparisons with strings and one of the operators <, <=, >=, > (to catch some instances of use of & rather than &&). 20090926 Changing default value of option show_boundtol from 0 to 1 and adding option boundtol_max (default 1e-5). When "option show_boundtol 1;" is in effect, report changes that would affect dual variable computations whenever dual variables are recomputed, not just when a solution is read, but report them only if the new $abs_boundtol or $rel_boundtol value would not be more than $boundtol_max. 20091014 In the MS Windows AMPL binaries, fix a glitch with "solve" and "shell" commands in handling explicit paths containing blanks, and fix a possible bug in determining the environment passed to solvers and shell processes. 20091101 Fix a glitch that (on some systems) caused this example: var x >= -900 <= -800 := -860; var y = x*exp(x); minimize zot: (x + 870)^2 + y; solve; to give error message Error differentiating =y: can't compute exp(-870). 20091112 Fix an erroneous bad-subscript message that arose under complicated conditions. In single-step mode, try to tolerate some syntax errors more gracefully when reading stdin (assumed to be a human typing). Exclude defined variables (and their defining constraints) from the count of constraints used in deciding whether a problem is small enough for the student version. 20091120 Fix bugs with "solexpand _con[n];" when _con[n] is a defining constraint for a "var ... = ..." declaration (of a defined variable). Linear terms were omitted and the constant term had the wrong sign. A fault was also possible when presolve had not removed any ordinary constraints. Example: var x; var y = 3*x + x^2; minimize zot: exp(y); solexpand _con[1]; 20091122 Fix a glitch that caused an error message of the form "invalid refct -2147483647 in opgen" after execution of a huge number of suitable commands. (In the example that led to this fix, the message arose at _cmdno 7123157887, which is well over 2^32.) 20091123 Fix a bug with suffixes on constraints with indexing {if ...}. The example that led to finding the bug produced an invalid .nl file on which the solver issued a "bad line ..." error message. 20100122 Fix glitch with printf's handling of negative numbers with %a or %A. The sign was ignored that the exponent was wrong. 20100226 Fix a bug in the expand command's printing of complementarity constraints (which erroneously included an Infinity). 20100228 Allow an optional output redirection after "commands filename", which applies to output that would otherwise go to the standard output. Closing the target of the redirection during execution of the commands in filename causes such output again to go to the standard output. 20100320 Fix a fault with "option presolve 0" on some problems with piece-wise linear terms. (Work-around: leave option presolve at its default setting.) 20100330 Fix a bug (fault) with "reset data;" when (e.g.) two params indexed by the same set were affected. Example: set S; param n_min{S}; param n_max{S}; table foo "nutrients.tab" IN: S <- [NUTR], n_min, n_max; read table foo; display n_min, n_max; reset data; # faulted 20100402 Fix a bug (possible fault or worse) in handling nonconvex piecewise-linear terms. Values for introduced variables that were then eliminated by substitution were mistakenly stored after "solve" with a possibly out-of-bounds subscript. (Whether this caused trouble was problem-dependent.) Work-around: "option pl_linearize 5;". Fix a bug with applying piecewise-linear operators to defined variables. Incorrect problem transforms sometimes resulted. 20100429 Fix a bug (fault) in handling valid input followed (e.g.) by an empty line and "*/" without a newline at the end of file. 20100705 Fix trouble that could arise in 32-bit binaries after they execute more than 2^31 commands. 20100708 Fix a bug introduced 20100228 in the "commands" command: when used inside a loop, a fault (or worse) was possible. 20100715 Fix a glitch in the regular expression functions (sub and gsub): ^ only matched the beginning of the string and not the beginnings of internal lines (i.e., the text following a \n character). Now it also matches the character following each \n (just as $ matches the end of each internal line as well as the end of the string). Example: before the bug fix, num0(sub(solve_message, '(.|\n)*\n([0-9]+) MIP(.|\n)*', '\2')) was the only way to extract the number of MIP simplex iterations from the solve_message returned by cplex. The bug fix lets us replace the first \n with ^: num0(sub(solve_message, '(.|\n)*^([0-9]+) MIP(.|\n)*', '\2')) Extension to the regular expressions processed by sub and gsub: @ is similar to ., but also matches \n, i.e., @ is a synonym for (.|\n). The above example can now be written num0(sub(solve_message, '@*^([0-9]+) MIP@*', '\1')) Of course, a literal @ appearing in a regular expression must now be escaped, as in sub('x@y', '\@', ' at ') in which previously the \ was not required (but was allowed). 20100915 Fix a bug (fault) with variables declared to be in a union of more than two intervals. Fix bugs (e.g, possible fault) in cleaning up after a solve with piecewise-linear terms or variables in unions of intervals, and with models that had both piecewise-linear terms and variables in unions of intervals. 20100928 Adjust environment changes so if one declares two problems, (with distinct environments) then sets option randseed to the same value in each, switching between these problems will not cause option randseed to be reset to the common value, but the random number sequence will continue. For example, var x; minimize zip: sin(x); minimize zap: cos(x); problem p1: x, zip; problem p2: x, zap; option randseed 42; print Uniform01(); problem p1; option randseed 42; print Uniform01(); problem p2; print Uniform01(); used to print "0.31376060558952107" thrice. Now the third "print" gives a different value (0.8653543517772044). Best practice (least confusing) is to adjust option randseed before declaring problems or new environments, in which case the this change to environment processing makes no difference. Fix a bug with _ampl_time and related quantities, such as _ampl_elapsed_time: they were only computed once and appeared to be constant thereafter. Fix a bug that could cause assignments to variable.relax not to take effect until completion of the outermost containing loop. (Insertion of "display somevariable.astatus;" after such assignments was a work-around; the choice of variable did not matter.) New builtin suffix "int" on variables is 0 for a continuous variable and 1 for an integer or binary variable (unless $relax_integrality is nonzero or the variable's .relax suffix is positive, in which case .int is 0 and solvers will see the variable as continuous). 20101103 Work around getenv() trouble seen on a version of Linux. The bug would cause a solver not to be found after a change to option PATH. 20101105 Fix a bug in converting decimal strings of more than 40 significant digits when the first appears after the decimal point. Example: print .010000000000000000057612911342378542997169; gave 4.008336720017946e-292 rather than 0.01. 20101125: Fix a bug (fault) with unsubscripted symbolic variables (a forthcoming feature, not yet described). 20101203: Fix a bug (fault, rarely seen) affecting 64-bit binaries. 20101206 Fix a bug (fault) possible with some uses of random variables (a forthcoming feature). 20110112 Fix a rarely seen bug (fault) that bit under the "right" conditions when making adjustments for new suffixes. 20110121 Fix a rarely seen parse error on a command after an in-line data section. Add warning that variables in subscripts are not yet allowed. 201102109 Fix a rarely seen bug that caused a fault in a complicated example. Fix a bug in "table write": symbolic params with numeric values were sometimes treated as having the corresponding string value. Add an error message for table declarations that involve row-index dummy variables in OUT or INOUT columns and do not specify "set -->" or "set <-->", as in the following example. set S; set T = 1..4; param p{S,T}; table goo: S <-- [r ~ R], {t in T} < p[r,t] ~ ("col" & t) >; Previously "write table goo" faulted (but worked fine with "S <--" changed to "S <-->" or "S -->"). 20110222 Fix a bug with solexpand applied to a defining constraint: a spurious "complements" sometimes appeared and a constant term was possibly lost. 20110306 Fix glitches with iterated columns in table declarations, i.e., with indexing < colg_1 [, colg_2, ...] > The indexing set was treated as the empty set in some situations, which might have caused "read table" not to read the columns in question or have caused "write table" to fault. Fix a glitch in the processing of "include filename" where filename is not quoted. If a nonblank character immediately followed filename, as in "include foo;" (in which semicolon is the nonblank character), that character was sometimes seen at the wrong time. For example, "if 1 < 0 then include f;" was treated as "if 1 < 0 then; include f". (Recall that an include phrase, such here as "include f", is processed before parsing; that is, "include" is unaware of the context in which it appears.) 20110321 Fix "sd botch in ordered_genall" in the following example: set S ordered by 1..10 default 1..10; display S; 20110413 Fix a fault that could arise in a complicated example with fixing and unfixing a defined variable within a bigger loop. Such fixing and unfixing is not allowed, but in the example, the bug bit before an error message about the inappropriate operation was issued. 20110419 Fix a fault in the "objective" command that was possible with some forms of iterated objectives. Example: set A dimen 2; var x; minimize Obj{(i,'N') in A}: x^i; data; set A := 1 x 2 N; objective; # faulted 20110524 Fix a bug (fault or worse) with some models having both defined variables and logical constraints. Fix a bug in simplifying logical constraints with constant logical expressions; the wrong truth value sometimes appeared in a platform- dependent way. 20110531 Adjust processing for $substout=1 to allow mutually recursive variable definitions from several indexed defining constraints. 20110614 Fix a glitch in computing Jacobian sparsity in some problems in which a nonlinear defined variable appears linearly in a nonlinear constraint and "option substout 1" is specified. Example: set I = 1..2; var x{I}; var y{I}; s.t. c1: y[1] = x[1] + 1; s.t. ydef: y[2] = y[1] + x[2] - x[1]; minimize zot: y[2]^2; s.t. bletch{i in I}: x[i]*y[i] >= 3; With $substout = 1, bletch[2] was incorrectly indicated to depend on x[1]. Aside from sparsity issues, this was harmless. 20110630 Omit "Highest address used = ..." from "Too much memory" message; the "highest address" seems pointless when virtual memory is used. Fix two bugs in handling complementarity constraints. In a command-line invocation of the form ampl -obfoo foo.mod foo.dat complementarities might sometimes have been misidentified, and "complementarities" of the form v complements expr1 == expr2 (which should just be simplified to expr1 == expr2) may have been conveyed as complementarities (which seemed not to bother the "path" solver). 20110713 Change the default computation of c.slack for a complementary constraint of the form c: L <= expr1 <= U complements expr2 with L and U finite to min(expr1 - L, U - expr1, if expr1 <= 0.5*(L+U) then expr2 else -expr2) Setting new option $old_complementarity_slack to 1 restores the old computation: min(expr1 - L, U - expr1, if expr1 <= L then expr2 else if expr1 >= U then -expr2 else -abs(expr2)) The new default computation is more useful than the old when an interior-point algorithm is used. With it, a slack value of zero means the constraint is satisfied exactly, a small positive value is a complementarity violation, and a negative value is a constraint violation. Unless we hear feedback that having option old_complementarity_slack is useful, this new option may eventually be withdrawn. When writing a .nl file after a solve or solution command, i.e., when primal or dual values for implicit variables or constraints are available, include their values, if nonzero, with the primal or dual initial guess. Fix glitches in "expand cc;" and "solexpand cc;", where cc is a complementarity constraint: bounds were sometimes not shown. Add two numbers to line 3 of the .nl header when complementarities are present: nd = number of "double" complementarities, i.e., those involving a double inequality, such as L <= expr <= U complements expr1 and nzlb = number of complemented variables with a nonzero lower bound. The new numbers are used in a new facility soon to be added to the AMPL/ solver interrface library for optionally modifying complementarity conditions to v1 >= 0 complements v2 >= 0. (With older versions of AMPL, the facility uses upper bounds on nd and nzlb.) 20110722 Fix some bugs with logical constraints. Problem commands did not cause logical constraints to be dropped or restored when they should have been. The "delete suffix ..." command ignored logical constraints. The "drop;" command did not show dropped logical constraints. The "delete lcon;" command, for lcon a logical constraint, misbehaved. For a logical constraint lcon, "expand lcon;" and "solexpand lcon;" sometimes printed an extraneous Infinity. 20110725 Fix a bug with redeclarations involving variables. Input of the form # ... write bfoo1; var newvar; redeclare minimize o: ... /* involving newvar */; write bfoo2; could have resulted in an incorrect foo2.nl. 20110813 Fix glitches in "show f;" when f is a pipe function. Fix a fault with empty slices that arose under rare conditions. In equality constraints in which one side involves numeric variables, assume the other side must also be numeric. This matters when the other side involves an if-then-else expressions whose "then" or "else" expression consists only of a dummy variable, such as set I := 1..2; var x{I}; var y{I}; subject to c{i in I}: x[i] = if i == 1 then y[i] else if i == 2 then i; which previously was treated as a logical constraint. (Changing "then i" to "then i + 0" was necessary to make constraint c into an algebraic constraint.) 20110825 Fix a parsing bug: an include phrase following an "if" command that could be followed by an "else" command (which might be in the included file), but is followed by something else was mishandled. 20110901 Fix a parsing bug ("syntax error") seen in the following example: param p; var x{1..2}; minimize zot: sum{i in 1..2} (1-x[i])^2; data goo.dat; if p > 3 then commands goo1.mod; expand; Changing "expand;" to "expand ;" or ";expand;" worked around the bug. 20110906 Extend "option show_stats 1" output to list the numbers (if nonzero) of equality, inequality, and range (i.e., double inequality) constraints. On Unix-like systems, add command-line option -g to start in a new process group. Extend option relax_integrality to apply to binary variables introduced to handle "in union_of_intervals" phrases in variable declarations. 20110908 Fix a bug with complementarity constraints. With certain (somewhat rare) problem dimensions, an incorrect .nl file was generated. An example where this bug bit: var x {1..2} >= 0; minimize obj: x[1] + x[2]; subj to con: x[1] <= 3 complements x[2] <= 5; 20110909 Fix a memory leak seen in a script involving a number of solves of large problems. Arrange for printing variables not to require instantiating constraints. With some command sequences, "ampl -t ..." or "ampl -T ..." or the equivalent "option times 1;" or "option gentimes 1;" will produce less output than before. 20110913 Fix a rarely seen bug that caused a fault in recent 32-bit Windows binaries. 20111005 Allow "ampl -R" (server mode) to use imported functions provided by "load" commands (but not pipe functions). 20111015 Fix a bug sometimes seen with "show" after a redeclaration: wrong variable names were sometimes reported (depending on details of the redeclaration). Example: var x; var y; minimize zap: x + y; s.t. C: 2*y <= 5; expand C; # OK param p default 2; redeclare s.t. C: p*y <= 5; expand C; # mentioned x rather than y 20111016 Fix more variable-reordering trouble with redeclare (not caught by yesterday's example). Example: var w; var x; var y; var z; minimize foo: w + 2*x + 3*y + 4*z; s.t. c1: x + 5*y <= 37; s.t. c2: y + 6*z >= 43; s.t. c3: 4 <= z + 7*x <= 19; expand; redeclare var y = x^2 + z; # "xrenumber bug" with version 20111015 expand; # wrong constraint c3 prior to 20111015 20111107 Supply missing initialization of SnprintF and VsnprintF in the AmplExports structure passed to imported functions and table handlers. (They have long been provided to imported functions used by solvers.) When imported functions or table handlers see ae->ASLdate >= 20111028 and ae->asl == 0, ae->Getenv(0) returns a char** value for the complete current environment. On Unix-like systems, set FD_CLOEXEC when opening files, to prevent child processes (e.g., from "solve" or "shell" commands) from seeing irrelevant open file descriptors, a bit of tidiness seldom relevant in practice. Fix the obscure glitch that prompts were preceded by # after reading a file that ended in an incomplete #comment line (one without a terminating newline character). The # persisted until the next complete comment had been read. Now the # should only appear if you manually enter an incomplete comment (e.g., under Unix or Linux, by typing control-D at the end of a # comment line). Fix a bug (e.g., fault) that bit under complicated conditions, e.g., with solving sequences of named problems. 20111110 Fix bugs (giving a surprising error message or fault) with updating sets over which subscripted objectives and problems are indexed. 20111121 Fix a bug with testing membership in "system sets", i.e., any of _CONS, _ENVS, _FUNCS, _LOGCONS, _OBJS, _PARS, _PROBS, _RANDVARS, _SETS, _SUFFIXES, _SYMVARS, _SYSTEM_SUFFIXES, _TABLES, _VARS or _VARSETS. The test sometimes came out wrong. Example: reset; print if 'XX' in _SETS then 'yes' else 'no'; # printed yes rather than no Add a warning about "data;" not being allowed in a compound command. Other error messages are likely to appear after the new warning. Note that "data filename;" and "data (filename_expr);" may appear in a compound statement. 20111216 For the -ix command-line option, in addition to allowing multiple files on separate lines of a suitably quoted x, allow multiple files on the same line if each file name is enclosed in single or double quotes. Add a brief explanation of multiple files to the "-i?" output. Fix a bug (introduced 20110531) with "option substout 1": an incorrect .nl file was sometimes generated when several constraints could be chosen to define a particular variable. Here is an example in which the bug bit, giving a .nl file incorrectly containing "nonlinear" constraints: option substout 1; var x >= 3; var y{i in 1..2} >= i; var z; var v = z + 3; var w = v + x; s.t. c{i in 1..2}: z = x + i*y[i]; minimize foo: x + z; 20120104 Fix a bug that could lead to an error message of the form "presolve has k = 11, P.nfc = 17". 20120117 Fix a glitch that caused the unload command not to fully unload the specified library. It's not clear whether this ever caused any visible symptoms other perhaps than some memory not being reclaimed. New builtin suffix ".sense_num" on objectives is -1 for minimization and 1 for maximization. New builtin suffix ".sense" is a symbolic version of ".sense_num" with default values "minimize" and "maximize" given by option sense_table. 20120126 Fix glitches with _obj.sense. Adjust "load" command to facilitate using a 32-bit AMPL binary Adjust "load" command to facilitate using a 32-bit AMPL binary with a 64-bit solver binary or vice versa: for a 64-bit AMPL, if the library name involves '.' and the final '.' is preceded by "_32", change the "32" to "64". Otherwise, if the library fails to load and there is a '.' in the name, insert "_64" before the final '.' unless it is already preceded by "_64". (For a 32-bit AMPL, the rules are similar, with the roles of "32" and "64" reversed.) The builtin set _LIBS still shows the names by which libraries were loaded, whereas option AMPLFUNC reflects the adjusted names. The unload command operates on names in _LIBS. Temporarily, at least, you can suppress the new behavior by specifying "option map_32_64 0;". Under MS Windows, add a test to the "load" command to ignore shared libraries (DLLs) compiled for the wrong number of address bits. (With more sensible operating systems, such libraries simply fail to load.) Fix a bug with option AMPLFUNC: after a sequence of the form load 'something'; load 'more'; unload 'something'; # or 'more' load 'another'; option AMPLFUNC did not show the new library. Change the default value for $ampl_libpath from "" to the directory containing the AMPL binary and (if different) the current directory. Absent command-line option -i (invoke ampl "-i?" for details), or an incoming $AMPLFUNC value, look for both amplfunc.dll and ampltabl.dll at startup by the AMPL search rules (rather, for ampltabl.dll, than by system-dependent rules). 20120131 Fix a possible fault during startup that could happen with version 20120126 on Unix-like systems. Fix a bug in simplifying a piecewise-linear term applied to a variable with a negative upper bound greater than the first negative breakpoint when the term has only two negative breakpoints and at least one positive one (after simplification for common slopes). Wrong information was transmitted to the solver. 20120208 Fix an obscure and rarely seen bug that arose under complicated conditions with repeated solves involving several entities having the same indexing when the size of the common index set changed. The example that revealed the bug faulted. 20120214 Correct the change of 20120208 to handle another case. On Unix-like systems, add command history processing similar to that of the "sw" program under MS Windows. The up- and down-arrow keys move among the history lines, and the left- and right-arrow keys move left or right one character in the current line. Where available, the "home" key moves to the start of the current line, the "end" key moves to the end of the current line, page-up moves up 10 lines and page-down moves down 10 lines. The control-right-arrow key moves forward one alphanumeric "word", and the control-left-arrow key moves left one "word". Control-W deletes the preceding "word". Control-D sends the current text without ending the line and signals end-of-file when there is no current text. Command history can be turned off by invoking "ampl -vi2 ..."; the -vi2 must come first; invoke ampl "-v?" for more on -v options. 20120217 Fix the bug that after, e.g., "display _VARS;", _VARS was not updated after subsequent variable declarations or deletions. 20120306 Add command-line option --version: when given as the sole command-line argument, --version causes AMPL to print its version and exit, regardless of any needed license file or manager. Adjust default computation of $ampl_libdirs to work correctly with non-ASCII directory names (where "ASCII" means "7-bit ASCII"). Note that AMPL has long handled UTF-8 encoding of Unicode, which can involve non-ASCII characters. On MS Windows systems, use of non-ASCII names can cause confusion when different code pages are involved. This is an issue outside the scope of AMPL. 20120317 Fix a bug (possible fault) in handling very long error messages. Fix a similar bug with long error messages in the ODBC table handler. Map NaNs in sets to a specific quiet NaN and fix a bug in testing whether NaN is a set member. Here is an example that now works and previously misbehaved: set S; param p{S}; data; param :S: p := NaN 37; print p[NaN]; 20120328 Fix a bug (possible fault) with variables and constraints indexed over a subscripted set. The fault was only possible if such a variable or constraint had to be regenerated. 20120406 Remove ampl_libpath from the default _LOCAL_OPTIONS (thus making $ampl_libpath available for possible use by the proxy table handle). Fix a glitch introduced 20120214 (on Unix-like systems) with a "shell;" command: the AMPL process could die from a SIGTTIN signal. 20120505 Fix a presolve bug with logical constraints (e.g., indicator constraints) that bit under complicated conditions. Fix a bug in the expand command's processing of a logical constraint that could cause it incorrectly to indicate whether presolve had removed the constraint. 20120507 Add a test to explicitly enforce an implicit limit of 199 on the lengths of names. 20120515 On MS Windows systems, if $TMPDIR has a nonblank value, use that value as the temporary directory name (as documented in the AMPL book). Otherwise use the value of $TEMP. Hitherto only $TEMP was considered on MS Windows systems. (On most MS Windows systems, $TMPDIR is not set, but $TEMP is set to a non-blank value by default; on such systems, this change should be invisible.) On MS Windows systems, fix a glitch with embedded blanks in the value of $TMPDIR or $TEMP (as appropriate). Blanks have long been tolerated in $TMPDIR on other systems. 20120516 Fix a bug with "redeclare" that gave error message "xrenumber bug" when the redeclared entity depended on something declared after something that depended on the entity. Example: param p default 6; param q = p + 1; param r = 2*q; display p, q, r; param s default 3; redeclare param q = p + s + 1; # "xrenumber bug" display q; 20120517 Fix a bug (fault) with some logical "atmost" expressions. 20120522 Fix a bug (giving "corrupted del_mblk arg") in "reset data p;" when p is a scalar param whose current value was computed from a default expression. The bug was exposed (previously harmless) was exposed by the changes of 20120214. Example: param p default 3; reset data p; # OK print p; # value computed from default expression reset data p; # "bug: corrupted del_mblk arg" On MS Windows systems, when process creation fails, report the path to the program that would not start, rather than just the program's name. 20120530 Fix a glitch under MS Windows with option debug_stub. Under MS Windows, if $Path does not appear in the incoming environment but $PATH does, use $PATH rather than $Path as the list of directories in which "solve" and "shell" commands should look for programs. Usually $Path is set, but at least some versions of the MinGW shell supply $PATH rather than $Path. 20120601 Extend warning that variables in subscripts are not yet allowed to most subscripts (not just subscripts on variables). 20120604 Fix two bugs in handling a problem with logical constraints that get converted to algebraic constraints: an incorrect .nl file was produced, and _logconname was miscomputed. 20120614 Fix a bug with piecewise-linear terms. When constraints added to linearize such terms increased the total number of constraints beyond the next power of two, a fault was sometimes possible. 20120619 Fix a glitch that gave rise to "bug: corrupted del_mblk arg" after a very large number of commands. 20120629 Fix another bug that could only bite after a very large number of commands. Note that using iterated commands, such as iterated "let", "printf", "drop", etc. commands, rather than "for" loops, where appropriate can greatly reduce the number of commands executed and give much faster executions. Using slices suitably sometimes also greatly speeds up execution. 20120804 Fix a glitch that gave error message "opnumber 243 in opgen" under complicated conditions. 20120831 Fix a glitch that gave rise to error message "sysset_en_IN called" in the following command: print if 'foo' in _AVAILFUNCS then 'yes' else 'no'; Add Addrandinit to struct AmplExports, which is available to imported functions in al->AE. An imported function or, more likely the funcadd_ASL function that makes known the imported functions in an imported-function library, can invoke addrandinit(rsi,v) (i.e., ae->Addrandinit(ae,rsi,v)) with apparent signature typedef void (*RandSeedSetter)(void*, unsigned long); void addrandinit(RandSeedSetter rsi, void*); to have rsi(v,$randseed) called initially with the current value of $randseed (as an unsigned long) and again whenever option randseed is assigned a value. This is meant to supply a new seed for random functions provided by the containing imported-function library. Like calls on at_reset() and unlike calls on at_exit, calls on addrandinit() are forgotten after a "reset;". Note that a "reset;" causes the funcadd_ASL() routines in all currently loaded imported function libraries to be called again. 20120911 Fix a bug in simplifying "and", "or", and comparison expressions when the result is constant (true or false). The example revealing the bug was of the form s.t. foo: x == 1 ==> 0 == 0; (with one of the zeros being an empty sum). 20121005 Use execve (on Unix-like systems) rather than execv when an invoked program appears to require interpretation by /bin/sh because the initial attempt to execute the program directly with execve failed. The idea is to supply the same environment in both cases. This change should mostly be invisible. 20121012 Fix a fault in simplifying logical expressions under complicated conditions. Example: var x{1..4} >= 1; s.t. fa: forall {i in 1..4} x[i] == 1 or x[4]^2 >= 1; s.t. b1: x[2] + x[3] <= 2; write 0; # fault 20121017 Fix a bug with a "solve" involving no logical constraints after a solve with one or more logical constraints. The bug could cause a fault or other surprising behavior. 20121021 Fix a bug (sometimes leading to a surprising "solve_out:..." error message and/or an invalid .nl file) with problems containing both logical constraints and "var ... in ..." declarations, such as var x{S} in {0, 4, 6}; 20121024 Fix a bug with "var ... in ... union integer [...,...];", as in var x in {2.5} union integer [3,5]; minimize zot: (x - 4.3)^2; The integer interval was treated as a continuous interval. Now, when appropriate, "var ... in integer [...,...] ..." will treat the variable as an integer variable. 20121116 When a defined variable involves piecewise-linear terms but is otherwise linear and when convexity (or concavity) permits all appearances of the defined variable in constraints and objectives to be linearized, do so. 20121120 Fix bugs (faults) with min() and max(), which now return Infinity and -Infinity, respectively. 20121210 Fix a possible minor glitch (irrelevant "_32" or "_64") in an error message about the failure of a "load" command. Fix potential trouble (not yet actually seen) with generic synonyms in 64-bit binaries. 20130109 Add an error message about a constraint, objective, or defined variable declaration that has a piecewise-linear expression with a variable in the slope or breakpoint list. In printing commands (display, print, printf), variables can appear in slope and breakpoint lists, but not in declarations. 20130117 Diagnose attempts to add "delete" or "purge" commands to a compound command (e.g., a for or repeat loop) via a "commands" command. Such attempts previously could lead to a fault. 20130201 Fix a bug (e.g., fault) with read commands having nested indexings in which an inner indexing involves an outer dummy variable. 20130207 Fix a bug with $ampl_libpath after a command sequence of the form # nontrivial declarations and/or commands reset; reset options; # option ampl_libpath might now have been bogus 20130218 Fix a bug in deducing reduced costs involving nonlinear if-then-else expressions whose "if" test has a "count" expression. Such expressions are unlikely in practice, but here is an example: var x := 3; minimize f: if count(sin(x) > .1, cos(x)>.1, tan(x) < -.1) >= 2 then x^2 else x^3; display x, f, x.rc; # x.rc was wrong let x := 3.1; display x, f, x.rc; # x.rc was accidentally right Fix a bug with default expressions for scalar params: if evaluation of the default expression failed, e.g., due to a missing value, the param subsequently appeared to have an unpredictable value. This was only an issue in an interactive session, such as ampl: param a; ampl: param b default a + 1; ampl: print b; Error executing "print" command: errror during evaluation of b: no value for a ampl: print b; 0 # a potentially random value Add a complaint about integer variables declared with bounds that when rounded (up for the lower bound, down for the upper bound) to integer values are inconsistent. Example: var x integer >= 0.1, <= 0.9; maximize obj: x; solve; display x; # Previously said "Solution determined by presolve;" # now says "Infeasible constraints determined by presolve." 20130221 Fix a bug, leading to "unexpected type ... in lsimplify()", with <==> in logical constraints. Example: var x{1..2}; s.t. c: x[1] == 1 <==> x[2] == 2; Fix bugs with logical expressions involving "not" (i.e., "!") and inequalities involving NaNs. For example, param p; data; param p := NaN; print if !(p > 3) then "yes" else "no"; # incorrectly printed "no" Similarly, in the (silly) model var x; s.t. c: 1 < 2 ==> !(x > 4); render c as the logical constraint !(x > 4) rather than the arithmetic constraint x <= 4. 20130222 Fix a bug with multiple t-headers in data sections, leading to error message "irregular : ... : header". Here is an example, previously rejected, of multiple t-headers for set and param data: set ORIG; set DEST; set PROD; set ROUTES within {ORIG,DEST,PROD}; param cost{ROUTES}; data; set ROUTES : FRA FRA DET DET LAN LAN WIN : bands coils bands coils bands coils bands := GARY - - - - - + - CLEV + + + + + + - PITT + - - - - - + : WIN STL STL FRE FRE LAF LAF : coils bands coils bands coils bands coils := GARY - - + - - - + CLEV + + + - - + - PITT - + - + + - - ; param cost: FRA DET LAN WIN STL FRE LAF : bands bands bands bands bands bands bands := CLEV 27 9 12 . 26 . 17 PITT 24 . . 13 28 99 . : FRA DET LAN WIN STL FRE LAF : coils coils coils coils coils coils coils := GARY . . 11 . 16 . 8 CLEV 23 8 10 9 21 . . PITT . . . . . 81 . ; set ORIG := GARY CLEV PITT ; set DEST := FRA DET LAN WIN STL FRE LAF ; set PROD := bands coils ; display cost; Note that there is a mistake on line 4 of p. 476 of the AMPL book: the colon before the first "t-header" should be omitted. 20130226 Fix a bug with fixing variables that appear in complementarity constraints when the fixing should cause the constraints to be removed by presolve. Example: var x{1..7}; s.t. c{i in 1..6}: x[i] >= 0 complements x[i+1] >= 0; fix{i in 1..3} x[i]; solve; # gave error message "presolve has k = 6, P.nfc = 7" 20130327 Fix a bug involving imported functions and defined variables, illustrated by the example: load amplgsl.dll; function gsl_cdf_ugaussian_P; param y default 0; var theta; var z = gsl_cdf_ugaussian_P(theta*y); display z; which said "z = 0" rather than "z = 0.5". The bug was that after 0 was substituted for theta*y (since y has its default value 0), gsl_cdf_ugaussian_P(0) was ignored. 20130408 Fix a bug with "option linelim 0;" that affected some complicated examples involving linear and nonlinear defined variables. The linear variables were sometimes evaluated after nonlinear variables that used them, leading to wrong values for the nonlinear defined variables. Note that "option linelim 0" can lead to substantially faster processing of some problems, but makes linear defined variables appears as nonlinear. For some nonlinear solvers, this makes no difference, but it cannot be used with purely linear solvers. Somewhat improve performance in handling many linear defined variables with the default "option linelim 1". In one example, time to invoke the solver was reduced by 16%. The "temporary" debugging interpretation of the "4" bit of $linelim, describe in the change-log entry of 20011206, is rescinded. (Version 20130407 had a glitch in this change.) 20130409 Fix a bug giving "unexpected type 0x960c in augment_incidence()" with nonlinear defined variables involving an if-then-else expression whose "if" test is constant. 20130417 Fix a bug "solve" involving complementarity constraints, followed by "fix" command(s) that remove all complementarity constraints, followed by a "solve": an incorrect .nl file was written for the second "solve". Fix an optimization bug with tests of the forms expr != numeric_expression expr == numeric_expression numeric_expression != expr numeric_expression == expr If expr was not numeric, the numeric_expression was not evaluated, so an error in it was not detected. Example: set S = {'abc'}; print{i in S, j in S: i != j+1}: i,j; printed "abc abc" rather than complaining that j+1 cannot be evaluated. Fix an obscure bug with complementarity constraints that led to an error message of the form "presolve has k = nnn, P.nfc = mmm". 20130422 Fix a bug (possible fault) with logical constraints when presolve eliminates all of them. 20130427 Fix a bug with logical constraints: on problems having both logical and algebraic constraints, when presolve eliminated all logical constraints but not all algebraic constraints, a fault was possible. Now, on other problems, "Solution determined by presolve" is more often possible. Fix a bug with "display _logcon;" on problems having both logical and arithmetic constraints: wrong values were displayed. The bug did not affect, e.g., "display {i in 1 .. _nlogcons} _logcon[i];". 20130501 Fix some bugs (infinite loop, fault, or surprising error message "nonvariable type ... in eput") in simplifying some complicated logical expressions. 20130502 Fix a bug with loop invariants having internal dummy variables. In the following example, the inner sum was moved to the wrong place, giving a fault: set I; set J; set S within {I,J} default {};; param p{i in I} = sum{z in {0}: sum{j in J: (i,j) in S} 1 > 0} 1; let I := 1..10; let J := I; display p; In this example, "sum{j in J: (i,j) in S} 1" is a bizarre way of writing what more clearly and efficiently would be written "card{(i,j) in S}"; making this change bypassed the bug, as did changing the inner sum to the clumsier "sum{(i,j) in S} 1". 20130506 Fix a glitch with error recovery: after an error message about "no value for ...", supplying the data and then trying to solve sometimes led to an invalid .nl file. Example (typed on stdin): set I = 1..2; var x{I} >= 0; param p{I}; s.t. c: sum{i in I} x[i] + sum{i in I} 2*x[i] <= 3; minimize zot: sum{i in I} p[i]; solve; # error processing objective zot: no value for p[1] data; param p := 1 -1 2 -2; solve; # bad line 22 of ....nl Fix a glitch giving an "unexpected nonvariable type..." error message when option presolve 0 is unwisely used with certain logical constraints involving "exists" or "forall". 20130510 Fix a glitch with _ncons, which incorrectly gave the number of algebraic constraints plus the number of logical constraints rather than just the number of algebraic constraints, i.e., _ncons was high by _nlogcons. Name Meaning _nlogcons number of logical constraints before presolve in the current problem _snlogcons number of logical constraints after presolve, i.e., as seen by the solver _logcon indexed by {1 .. _nlogcons}: true (1) or false (0) values of logical constraints before presolve _slogcon indexed by {1 .. _snlogcons}: true or false values (1 or 0) of logical constraints seen by the solver _logconname indexed by {1 .. _nlogcons}: names of logical constraints in the current problem _slogconname indexed by {1 .. _snlogcons}: names of logical constraints seen by the solver 20130530 Add an error message that symbolic variables are not yet available. Fix a fault with deleting check commands. Example: param p; check p > 3; delete check 1; # faulted 20130608 Fix a glitch (fault after 2 "reset;" invocations after the relevant "load" command) with imported libraries that make no addfunc(...) calls. 20130621 Adjust "load", "unload", and "reload" commands to canonicalize library names by simplifying paths and, for MS Windows, changing / to \. For example, "a/b/../c/foo.dll" becomes "a/c/foo.dll" and "./foo.dll" becomes "foo.dll" in _LIBS and the associated test for whether the library is already loaded. Nonetheless, "./foo.dll" is sought only in the current directory, whereas "foo.dll" is sought in the directories listed in $ampl_libpath. It is still possible to load the same library more than once by using a different name for it, either via a hard or symbolic link, by using such variations as "foo.dll", "foo_32.dll", "foo_64.dll", or using a name that starts with "../" or otherwise involves enough instances of "/../" to take the apparent name above the current directory. Now _LIBS reflects the canonicalized names, and possibly simplified full pathnames appear in $AMPLFUNC. Fix some glitches under MS Windows that, depending on the compiler used, could lead to a "Cannot find" rather than a "Cannot load" error message. 20130624 Fix more trouble with "load" that may only matter for MS Windows. 20130625 Fix trouble introduced 20130624 with the incoming $AMPLFUNC. 20130704 Adjustments for handling problems with >= 2^31 Jacobian nonzeros. New .nl file format codes z, h, and (temporarily, for debugging) B, G, H, and Z. The z format has a "K" section instead of a "k" section; the K section has Jacobian column lengths rather than the the k section's sums of such lengths, permitting entries in the K section to be encoded with 32-bit integers as long as the problem has less than 2^31 variables. (Like the b format, the z format is a binary format that only uses 32-bit integers. The z format uses 2 bytes per operation code rather than 4, so it is somewhat more compact than the b format.) When the b format is requested and the problem has at least 2^31 Jacobian nonzeros, AMPL now uses the z format instead. The h format is another binary format, one using 64-bit integers, permitting encoding of problems with more than 2^31 variables or constraints. The current AMPL cannot process such large problems, but a later version should be able to do so, at least when appropriately compiled. The h format is available now to permit preparing solvers that can handle it. Debugging format G is like g, except that it has a K section. Debugging formats B, H, and Z are like b, h, and z, but with reversed byte encoding of binary numbers. (The solver interface library adjusts for byte ordering if necessary, allowing one to generate a binary .nl file on a big-endian machine and solve it on a little-endian machine or vice versa.) Reading the new formats requires use of versions >= 20130703 of ASL, the AMPL/solver interface library. 20130731 Fix a bug with inserting "_64" or substituting "_64" for "_32" in file names given to "load" commands with the 64-bit MS Windows binary and vice versa with the 32-bit MS Windows binary. 20130809 Fix a parsing glitch with entities named "end". In some contexts, such as "redeclare", "read table", "reset", "show", "suffix", "update", and "write table", "end" was treated as an "end" command. Fix a bug (fault) that was possible under some conditions with a set of variables indexed over the empty set. Fix a bug (wrong values; possible fault) in writing suffix values to .nl files for complementarity constraints involving expressions of more than one variable on both sides of "complements". The .nl file has two constraints for each such complementarity constraint, and both get the same suffix value. 20130815 Adjust 64-bit MS Windows binary to avoid trouble (not yet seen) with "sw" if ever a thread handle cannot be represented in 32 bits. 20130816 Fix a bug (fault) with several entities (e.g., constraints) using, using iterated quantities, such as sums, over the same subscripted set with a constant subscript. 20130903 Fix a fault that was possible with some problem instances involving piecewise-linear terms. 20130906 Fix a bug that kept suffixes on logical constraints from being transmitted to the .nl file. 20130921 Fix a possible fault (or worse) with instances having piecewise-linear terms whose linearization caused the number of variables to grow beyond the next power of 2. Fix a bug unlikely to bother many people: a fault with duplicate false constraints: c: 0; d: 0; display c; # faulted 20130926 Fix a bug (eventual fault or worse) with an increase (above the next higher power of 2) in cardinality of the indexing set of a defined variable, some of whose values had been used in printing or other commands. 20131002 Fix a glitch with "ampl -b4 ..." in the error message for an unavailable solver: part of the error message did not appear. 20131010 Fix a bug with two or more entities indexed over a subscripted set with constant subscripts in the entity declarations. The first-used subscript might have been used instead of the correct subscript in subsequent instantiations of the affected entities. Have "write ...", including "write 0;", set solve_result_num = 299 if presolve finds the problem infeasible. Fix a rarely seen bug introduced with the changes of 20130408 for option linelim 0. A faulty .nl file was sometimes produced, with defined variables incorrectly ordered. Defined variables may sometimes now appear in a different sequence in the .nl file. 20131012 Fix a bug that might have bitten after a "problem" command and before the next "solve" or "solution" command. If the "problem" command changed the constraints to be enforced or variables allowed to vary, if previous commands had caused presolve to run, and if nothing else had caused the need to run presolve again, commands such as printing and "let" commands that involved entities that depend on presolve might have used wrong values of some of those quantities. 20131018 Fix a bug with inserting "_32" or "_64" (as appropriate) before the final "." in filenames seen by load commands when the portion of the name before the "." was only one or two characters long (such as "cp.dll"). Alternatively, if a given library name has "_32" or "_64" before the final "." and fails to load (perhaps after changing "32" to "64" or vice versa, as appropriate), try omitting the "_32" or "_64". When a solution is read by "solve" or "solution" and no dual variables are present but could have been, and when presolve has removed some constraints, delay the computation of dual variables for the removed constraints until something, such as a request to print dual values, requires them to be computed from the then current primal and dual variable values (possibly left over from a previous "solve" or still at their initial values). 20131108 Fix a fault with redeclaring recursive declarations. Allow "display X;" when X is a param or set indexed over an infinite set and X has a recursive default value. The display just shows the values that so far had to be computed. Example: param f{i in Integers} default if i <= 1 then 1 else i*f[i-1]; set S{i in Integers} ordered default if i <= 1 then {1} else S[i-1] union {i*last(S[i-1])}; display f[6], S[6], f, S; 20131122 Fix a bug with "redeclare" that with some commands caused a delay in computing the redeclared value. Example: var x := 1; var y := 2; var z = y^2 + 1; display x, y, z; redeclare var z = 3*x + 4*y^2; display x, y, z; # z was wrong display z; # z was now right 20131203 Fix a fault that was sometimes possible with named problems. 20131212 Fix an infinite loop that was possible in interactive AMPL sessions after an error. Fix a bug (e.g., fault) that was possible with some instances after "option presolve 0;" following a "solve;" of a larger problem. 20131213 Allow execution to continue after "write 0;" gives an error message about infeasibility. 20140205 Fix bugs in simplifying logical constraints. The bugs could give an error message about "unexpected nonvariable type" or could result in a .nl file that could cause some solvers to complain about an invalid indicator constraint. 20140220 Fix a bug (error or fault) with "s.t. e: 0; display e.val;". (This was reported as corrected on 20140124, but the fix did not make it into the distributed binaries.) Improve simplification of piecewise-linear terms: after summing terms applied to the same variable, combine adjacent pieces with the same slope and turn the term into a linear term if only one slope remains. Example: var x; minimize zot: <<1,2;-1,2,-3>>x + sin(x) + <<1,2;1,-2,3>>x; solexpand; # now gives sin(x) rather than something messier Add new processing for defined variables whose value is a piecewise- linear expression on a single variable. Now the example var x; var y = 0.5*x + <<1,2,4;-3,1,-1,2>> x + .7; var z = <<-2.3, -1.55, -.8, 1.7, 3.7; -5, 4, -3, 2, -1, 6>> y; minimize zot: 2*y; minimize zot2: 3*z; gives the same .nl file (but for roundoff) as var x; minimize zot: 2*(0.5*x + <<1,2,4;-3,1,-1,2>> x + .7); minimize zot2: 3*<<-1.2,-.4,0.6,0.9,1,7/6,5/3,2,3,4,4.2,5.2,6; -15, 2.5, -5, 7.5, -10, 6, -4.5, 3, -1, 1.5, -7.5, 5, -2.5, 15>> x + 4.2; For now, at least, the "16" bit of $pl_linearize suppresses the new processing, i.e., "option pl_linearize 17" gives the former default behavior, and the "32" bit of $pl_linearize causes explicit definitions of variables involved in nonconvex (or nonconcave, as appropriate) piecewise-linear terms when this gives a sparser representation. Thus "option pl_linearize 33" may sometimes result in the output of "solexpand;" being slightly easier to read. Today's changes include recording more pl_linearize bits, including 4 and 8, which may cause changes to .sol files when nondefault $presolve_fixeps values matter. 20140224 Fix a bug with defined variables in tables: wrong values may have been transmitted by "write table". Explicitly subscripted defined variables and those with a ".val" suffix were not affected by the bug. Fix a bug (fault) with "solexpand _con;". 20140312 Fix a bug (fault) that crept into version 20140220. Fix a bug with "read" commands in compound statements (e.g., loops) that are executed after a "solve" or "solution" command that introduces new suffixes. New option cmdtrace: "option cmdtrace 1;" turns on printing of the name of each command executed and, when the command comes from a file other than stdin, the filename and line number of the command. 20140320 Fix a bug with problems with general complementarity conditions (expr1 >= 0 complements expr2 >= 0 where neither expression is linear in just one variable) and some partially linear nonlinear defined variables. The example that revealed the bug resulted in an invalid .nl file. 20140328 Fix a bug with sets that appear in loops and involve defined variables, illustrated by the following example. set K; var x {K}; var y {k in K} = x[k]; data; set K := 1 2; var x := 1 1 2 0; for{i in 1..2}{ display x, y; display {k in K : x[k] == 1}; display {k in K : y[k] == 1}; #was wrong in 2nd iter let x[2] := i; } Fix a bug with problem declarations involving iterated entities. Changes to things that indirectly affected the sets over which the entities were iterated did not cause the drop/restore state to be updated. Fix a fault that was possible in interactive sessions with "show x;" after an error message about the wrong number of indices in the declaration of x. 20140330 Work around Microsoft bugs that caused Infinity^2 to give 0. 20140331 Fix a bug (causing an error message "Can't evaluate _con[...]") with "expand _con;" when logical constraints are present. Use "expand _logcon" to see the logical constraints before presolve. 20140409 Fix a glitch whereby some commands, such as "fix", that should not require the current problem to be fully instantiated, nonetheless did so, possibly leading to an error message about "no value" for sets or parameters involved with variables, constraints, or objectives not involved with the current command. 20140505 Fix a bug ("corrupted del_mblk arg") with some sequences of solves of subscripted problems. 20140513 Fix a bug that could corrupt memory after a "solve" when "let" assigns to a variable that is not in the current problem but was declared before a variable that is in the current problem. Fix a bug with sequences of commands in which the first of two sets that previously had the same value is assigned another value. In some cases a computed value changed when it should not have been affected. Example: set A; set B; set C; set D = A union C; set E = B union C; data; set A := a b c; set B := a b c; set C := x y; display D, E; update data A; data; set A := p q r; display D, E; # E was wrong 20140524 Extend test for "no variables, but lower bound = ..." to fix a glitch with "option presolve 0" (which is often a bad idea) whereby infeasible constraints with no variables were eliminated when defined variables were present. When no defined variables are present, such infeasible constraints are passed to the solver. 20140630 Adjust computation of _ampl_elapsed_time, _shell_elapsed_time, _solve_elapsed time to reduce roundoff error and, for MS Windows, to report elapsed time rather than CPU time. Fix glitches with adjustments for some complementarity constraints that could lead to incorrect .nl files with some uses of nonlinear defined variables or to incorrectly reading dual variables from a .sol file. Fix a bug with _con[i].status that could give "log" instead of "sub". New option display_set_1col is analogous to $display_1col, but applies to sets: if S is a set with $display_set_1col >= card(S), the "display S;" lists each element of S on a separate line. The default value 0 for $display_set_1col causes no changes to the hitherto seen behavior of "display S;". 20140704 Fix a fault after the error message "nonconstant left-most expression in double inequality constraint". 20140711 Fix a glitch on 64-bit MS Windows whereby printf "%x\n", 0xabcd1234; printed "ffffffffabcd1234" rather than "abcd1234". 20140827 Fix a bug revealed by the following example: var x := 3; var y; s.t. ydef: y = x^2 + 1; minimize zot: (y-4)^2; print zot; # faulted 20140804 Fix a bug with "option substout 1;", which did not detect and cope with circular dependencies. Example: option substout 1; var x; var y = x^2 + 1; s.t. zap: x = y; display y; # printed 0 rather than 1 20140908 Fix a rarely seen fault (or unpredictable behavior) and a bug that gave rise to the error message "named_genall called". 20140923 Fix a possible fault with # various declarations minimize f: /* nonlinear expression */ # more declarations and data display f; solve; 20140925 Fix an fault that arose under unusual and complicated conditions. 20140926 Fix a possible fault with "let" commands assigning to synonyms, e.g., let{i in 1 .. _nvars: _var[i] <= 0} _var[i] := .1; 20141016 Fix a bug with updating dependencies for commands when "solve" or "solution" command within a loop causes a new suffix to be introduced. In the example that revealed the bug, a command of the form "reset data foo;" appeared to be ignored after the new suffix appeared. On Unix-like systems, arrange to catch and report surprise SIGPIPE signals. 20141020 Redo bug fix with updating dependencies for commands when "solve" or "solution" command within a loop causes a new suffix to be introduced to make the fix work in more cases. 20141024 Fix a glitch with "solexpand": when some but not all members of an array of constraints were dropped, "solexpand" sometimes incorrectly commented that constraints were eliminated by presolve. Do not regard "q" as "quit" if followed on the same line by other text, other than ";". Fix a glitch with "not alldiff{...}...". 20141030 Fix a "yacc stack overflow" that was possible with unusual models. 20141104 Fix a bug that gave a surprising "presolve has k = ..., P.nfv = ..." message instead of "Solution determined by presolve; ...". 20141117 Tweak error message for failed evaluations an imported functions to mention "gradient" or "Hessian" if appropriate (i.e., if Errmsg starts with ' or ", in which case the initial ' or " is not printed). Fix a memory leak with for{...}{...} loops. Fix a bug with new suffixes returned by solvers: if subsequent commands increased the number of components of the suffixed variable, constraint, or objective beyond the next power of 2, memory corruption could result. In the example behind this bug fix, memory corruption caused the error message "bug: corrupted del mblk arg". 20141124 Fix a bug with expressions involving values of dropped objectives or of constraint bodies of dropped constraints when the expressions appear in commands, such as printing commands, and previous commands, such as "let" commands, have changed something involving variables involved in the expressions. Under some conditions -- without a "solve", "solution", or "write" command, including "write 0;", being executed after the changes -- the expressions were not computed correctly. Fix a bug with "solve; display v;" when v is a defined variable. Under some conditions, v was not recomputed to reflect the results of the solve. Fix two bugs with redirecting a printing command (display, print, or printf) to a file that cannot be created. The next command could elicit a surprising error messages, such as "... contexts remain open". 20141128 On nonlinear problems, defer computation of dual variables for constraints eliminated by presolve until they are needed for some reason (such as display), lest inability to compute derivatives of nonlinear expressions should give an error message and terminate the run. 20141206 Fix a couple of small memory leaks that accumulate over many commands. Fix a bug evaluating dropped objectives that depend on defined variables. Example: var x := 2; minimize foo: (x-3)^2; var xi{i in 1..3} = x^i; minimize o{i in 1..3}: xi[i]; problem p: x, foo; display o, xi; # o was shown as all zeros. 20141216 Fix a bug introduced 20141206 with defined variables not appearing in the current problem (i.e., not in any constraint or objective): such a variable was sometimes miscomputed. Example: param p default 0; set I = 0..1; var x{i in I} >= 0 := i; s.t. c: sum{i in I} x[i] = 1; var y = x[0]; var z = p*y; print z; # printed 1 rather than 0 20141228 Fix a bug introduced in version 20140220 that sometimes bit with defined variables involving piecewise-linear terms, leading to incorrect results. 20141231 To work around a Microsoft bug, set $ampl_funclibs to "-" rather than "" when "solve" does not involve imported functions. 20150127 Fix a glitch with format %.0e that caused printf "%.e\n", 4.7; to print "4e+00" rather than "5e+00". Fix a fault with "show table_name;" when table table_name involves nested indexings. Example: set S; set T; param p{S, T}; table mytab : [S], {j in T}<{i in S} p[i,j]~(i&j)>; show table mytab; # faulted 20150131 Fix a possible fault with loops involving indexed collections of objectives and increases to the cardinality of the indexing sets of those objectives. 20150213 Extend simplification of logical constraints so a constraint of the form logical_condition ==> linear_constraint in which logical_condition is found always to be true is reduced in more cases to a linear rather than a nonlinear constraint. Fix a glitch in the reported number of eliminated algebraic constraints in the output of "option show_stats 1" when logical constraints are converted to algebraic constraints. 20150214 Fix a bug that bit under complicated conditions and gave error message "tva top error". 20150302 Fix a bug in assigning suffixes to subscripted problems by "solve" and "solution" commands. Suffixes were assigned to the problem's first subscript rather than the current one. 20150312 Fix a glitch in processing UTF8-encoded Unicode that caused declared names starting with a non-ASCII character to be mishandled. 20150313 Adjust printf format %q to treat non-ASCII Unicode characters as alphabetic characters. This affects the display command's printing of set members. 20150325 Fix a bug (possible infinite loop) with a constraint of the form "1 ==> algebraic_constraint" when the algebraic_constraint involved a complicated sum. 20150327 Fix a bug (possible fault) that could bite during cleanup after an error message arising during instantiation of a defined variable. 20150415 Fix some bugs that could affect models with logical constraints, defined variables, and piecewise-linear terms, and fix a bug with "option presolve 0" applied to such models. 20150422 When history is desired (e.g., stdin is not a file) but cannot be instantiated, allow "solve" and "shell" to work without the need to invoke "ampl -vi2". This is only known to be relevant to 32-bit Alpine systems. 20150430 Fix a bug with logical constraints that bit under complicated conditions. 20150501 Redo yesterday's bug fix. 20150505 Fix some glitches with error processing. 20150516 Fix a glitch with v.lb, v.ub, v.lslack, v.uslack, and v.slack where v is a variable instantiated without need of presolve and after one or more other variables have been instantiated. Example: var x <= 0; var y <= 0; display y.lb; display x.ub; # x.ub was wrong (with separate display commands) # but all went well with "display y.lb, x.ub;" 20150630 Fix a glitch with output redirections, which could get lost during computation of duals for eliminated constraints. 20150721 Fix bugs with defined variables that are used as the only variable in an equality constraint. Solvers saw correct problems, but commands (such as display and print) that used expressions involving the defined variables saw incorrect values for those variables. Example: var x := 2; var y = (if x >= 1 then (x-1)^2) + 2.5; s.t. c: y = 17.5; display x, y, y.defeqn; # gave y = -16 rather than 3.5 # and y.defeqn = 0 rather than 1 Fix a bug in the indexarity functions: scalar variables gave -1 rather than 0. Example: var x; print indexarity('x'); # printed -1 rather than 0 20150815 Fix glitches in detecting domain or range errors in "math" functions such as log and exp on some systems, mainly MacOSX and hpux-ia64. 20150827 Under "ampl -b4", have error messages show context. 20150910 Fix a bug with deducing that a defined variable has a fixed value. If the number of variables was sufficiently more than the number of constraints, memory corruption was possible. 20150923 Fix a bug with defined variables in scripts that change the cardinality of indexing sets of variables declared before the defined variables. A fault or other wrong behavior was possible. 20150925 Fix some trouble (possible fault) with the update of 20150923. 20150928 Fix another bug similar to that in the update of 20150923. 20151010 Fix bugs with some complicated uses of defined variables and piecewise-linear terms. Examples: var x; var y = x + 3; s.t. c: y = 7; minimize o: <<4;-2, 3>> x; write 0; display x, y, o; # printed "y = 3" rather than "y = 7" and (illustrating a separate bug) node n{1..3}; arc a{i in 0..2} from {if i >= 1} n[i] to {if i < 2} n[i+1]; s.t. c: a[1] = 3; var x = a[2] + 1; var y; minimize o: x + <<7;-1,1>> y; solve; display o, x, y; # o and y were 11 and -7 rather than -3 and 7 20151015 Fix a bug (possible fault) with dropping all constraints and objectives that use a defined variable. Fix a bug with evaluating piecewise-linear terms in problems involving defined variables fixed by presolve. A wrong value might have been computed, e.g., for use in display. 20151026 Fix bugs that caused faults under complicated conditions. 20151029 Fix some trouble updating variables under complicated conditions. Fix a glitch with "xref": named problems were listed twice. 20151104 Fix bugs with problems involving complementarity constraints and defined variables. Under complicated conditions, incorrect .nl files were written. Complain in more cases about objectives and constraints involving NaNs or bounds with inappropriately signed Infinity. (There was already a complaint about linear expressions with a Nan or Infinity as a coefficient.) 20151118 Fix a bug with "expand c;" when c is a constraint of the form double-inequality complements expression or expression complements double-inequality. If the expression involved a constant term, the constant term did not appear in the "expand" output. Example: var x{1..3}; s.t. c: 3 <= x[1] <= 5 complements 123 + x[2] + 2*x[3]; expand c; # 123 did not appear; it did appear in "solexpand c;" 20151120 Fix a bug with introduced 20151104 in handling complementarity constraints involving an explicit upper bound of +Infinity. An incorrect error message about the Infinity was given. 20151130 Fix a bug with "read table" or "write table" that gave error message "1 contexts remain open after compile!" under complicated conditions. Disallow defined variables in IN or INOUT table declarations. 20151203 Fix a bug with computing values of defined variables eliminated by presolve: the linear parts were ignored. 20151216 Fix a bug with using variables when params could be used: an assignment to an otherwise unused variable of a value involving a defined variable used in the problem just solved generally used 0 as the value of the defined variable. Example: var F; # could be param F, as solvers do not see it set I = 1..2; var x{I}; var f = sum{i in I} (x[i] - i)^2; minimize o: f; s.t. vex: sum{i in I} x[i] = 1; solve; display f; # f = 2 (correct value) let F := f; display F; # F was zero 20151222 Fix a bug with reading large data sections (more than 16 or 32 kilobytes of values and subscripts for 32- and 64-bit binaries, respectively): data corruption or a fault was sometimes possible. 20160119 Arrange for defined variables that do not affect constraints not to participate in presolve deductions. Under unusual conditions, such defined variables could cause incorrect deductions. When more than 64 output files would be open, close the least recently used file and reopen it when necessary. 20160121 Fix a possible fault with "display foo;" when foo has a named 2-dimensional indexing set that is also the indexing set of an indexed collection of ordered sets. Fix possible trouble in the update of 20160119 during "reset;" when option log_file is not "". 20160131 Fix a rarely seen bug in changing the size of the indexing set of an indexed variable. Fix some bugs in the changes of 20160119. Fix a bug with "reset;" after "option log_file" specifies an invalid file. 20160201 Fix a fault that arose under complicated conditions. Fix trouble with "option log_File" when the solver faults. 20160207 Fix a fault with an invalid subscript message occurring after generation of a newly instantiationed variable declared before previously instantiated variables (an obscure situation). 20160211 Adjust debugging option nl_permute so when permutations are suppressed, the .nl file will indicate numbers of nonlinear constraints and variables based on the last nonlinear constraint or variable. Give a Caution message when suppressing permutations affects the .nl file. The "8" and "16" bits of $nl_permute can suppress these new Cautions. $nl_permute is now the sum of 1 ==> reorder constraints 2 ==> reorder variables 4 ==> reorder objectives 8 ==> suppress Caution about constraints 16 ==> suppress Caution about variables The default value for $nl_permute remains 3. Option nl_permute is not meant for use with solvers; great confusion may arise if binary or integer variables are present. Most users should avoid fiddling with $nl_permute. 20160221 Fix a bug (possible fault) with the interaction of blockmode (used with some API variants), option log_file, and shell or solve commands. New builtin function date(fmt,t) returns a string representing time t (as returned by the builtin time() function), formatted according to fmt (a character string). Either or both t and fmt may be omitted; time() is assumed if t is omitted and "%Y%m%d %H:%M:%S" is assumed if fmt is omitted; otherwise, date(t,fmt) and date(fmt,t) are treated alike. The expected usual usage is simply date(), i.e., omitting both arguments, as in print date(); to print the current time and date. The fmt string is similar to that allowed for the "date" command on many Unix-like systems, such as Linux and AIX; fmt is interpreted by the widely available strftime() library function. In short, various letters, when preceded by a % character, are replaced by details derived from the time t. Some, indicated by * below, may be affected by the current locale. Commonly available are %a abbreviated day of the week* %A day of the week, fully spelled out* %b abbreviated month* %B month, fully spelled out* %c date and time as preferred in the current locale* %d two-digit day of the month (1-31) %H 2-digit hour (00-23) based on 24-hour clock %I 2-digit hour (00-11) based on 12-hour clock %j day of year (001-366) %m 2-digit month (01-12) %M 2-digit minute (00-59) %p AM or PM %U week number of the current year (00-53), with week 01 starting on the first Sunday of the year %w day of the week (0-6), with Sunday = 0 %W week number of the current year (00-53), with week 01 starting on the first Monday of the year %x preferred date representation* %X preferred time representation (without date)* %y 2-digit year (no century) %Y 4-digit year %Z time zone %% % character Also available on some systems are %C similar to %c, but also including the timezone %D %m/%d/%y %e like %d, but with a space in place of a leading 0 %F %Y-%m-%d %G ISO 8601 4-digit year %g IS0 8601 2-digit year -- no century digits %h %b %k like %H, but with leading zero changed to blank %l like %I, but with leading zero changed to blank %n newline character %P lower-case %p: am or pm %s number of seconds since 19700101 00:00:00 UTC %t tab character %T %H:%M:%S %u decimal day of the week (1-7), with Monday = 1 %V ISO 8601 week number (01-53) %z the numeric time zone: +-hhmm (4 digits) On some systems, a #, E, or O may be inserted between % and some of the function letters indicated above to modify the conversion. Please experiment to see what works on your system. The fmt string may begin with %K or %L, which are removed and affect whether the rest of the fmt string reflects UTC time, formerly known as Greenwich Mean Time (%K), or local time (%L). Local time is the default, which %L simply confirms. If nothing remains after an initial %K or %L is removed, the default fmt is assumed. Thus date("%K") shows the current UTC time. The builtin ctime() function is extended to behave like date(), but with fmt = "%a %b %d %H:%M:%S %Y" assumed if fmt is omitted or is empty after an initial %K or %L is removed. Thus ctime("%K") shows the current UTC time, but formatted differently than date("%K"). 20160310 Fix glitches with floating-point values in hexadecimal notation (use of which is supported but very unlikely): some values that should overflow to (appropriately signed) Infinity, such as 0x1p1025, were mishandled, and values greater than 0x1p-1075 and less than 0x1.0000000000001p-1075 where treated as zero rather than the smallest denormal number. Fix a bug with with the current environment: after an environ command changed the environment, a problem command specifying the current problem did not change the environment back to the problem's environment. 20160325 When a shell command gives shell_exitcode > abs($shell_exitcode_max), resulting in "", show an option command that would have suppressed the "". The new printing of an option command is suppressed if $shell_exitcode_max < 0. Previously the test for issuing a "" was whether shell_exitcode > $shell_exitcode_max. 20160513 Fix a bug introduced in 20150721 that caused some defined variables involving linear uses of other defined variables to be miscomputed in the AMPL session (but not by solvers). Example: var w >=0.5 <=1 :=1; var x = sum{i in 0..4} w^i; var y = x; # linear use of defined var x var z = exp(y/w); # = exp(x/w) worked correctly maximize o: z; display x, y, z, objective; # z and o were wrong 20160519 Fix a bug (possible fault) in updating indexed symbolic params via "read table". Adjust ordering of sections in .nl files so bounds on variables and constraint bodies and primal and dual initial guesses precede constraint bodies. This change should be invisible to solvers that use the AMPL/solver interface library. For other solvers, turning on the "32" bit of $nl_permute will restore the old ordering. This interpretation of the "32" bit will eventually be removed unless we hear of a good reason to retain it. 20160530 Fix a bug in reading .sol files involving partially dropped logical constraints: memory corruption was possible. Extend option infeas_clear so "option infeas_clear 3;" allows "write" and "solve" commands to proceed when presolve detects infeasibility. In short, possible values for infeas_clear are now 0 ==> always suppress "solve" and "write" 1 ==> allow a second "solve" or "write" to proceed when reading stdin in interactive mode (default) 2 ==> always allow a second "solve" or "write" to proceed 3 ==> always allow "solve" or "write" to proceed Suppress some unlikely option commands issued while reading .sol files when they have no effect. 20160602 Fix a bug with some suffixes (of length 4-7) on synonyms. When the bug bit, memory corruption was possible. 20160609, 20160614 Fix a rarely seen bug in which the right-hand side of a "let" command involved a defined variable whose value was miscomputed. 20160712 Fix a bug with models having both "var in set" declarations and logical constraints. Incorrect .nl files or faults with "solexpand" were sometimes possible. Example: var x{1..2} integer in {11, 22}; s.t. c: alldiff{i in 1..2} x[i]; solexpand; #fault 20060714 Fix a bug with constraints, objectives, and defined variables involving a nonlinear expression of the form sum{i in S} if ... then i else ... or sum{i in S} if ... then ... else i in which "if ..." involved variables. An incorrect "if" operator could appear in the .nl file, which on some systems might lead to incorrect nonlinear evaluations in solvers. (On at least one system, a 32-bit Linux system, the bug was invisible.) 20160803 Change the message "xxx is not defined" from a warning to a fatal error (which terminates execution). 20160829 Fix a parsing glitch: do a better job of treating comments delimited by /* and */ as white space. Example: print if 1 < 2 then 3/*comment*/ else 4; # erroneously complained "3else is not defined"; now prints 3 20160907 Fix a glitch in the update of 20160829, which led to an inappropriate "syntax error" under complicated conditions. 20160920 Rather than always terminating execution after an error message of the form "nnn is not defined", when this message appears after a "model", "data", or "commands" command or include phrase issued in interactive mode, return to interactive input (after ignoring the rest of the offending input file). 20161005 Fix a bug that could bite after some error messages, such as (the unlikely) "cannot multiply ... by Infinity". No examples of the bug actually biting are known. 20161025 Adjust presolve to eliminate some defined variables that were previously rendered as constant-valued variables. Example: var x{i in 1..2} := i+1; var y = 1 / x[2]; # appeared in the .nl file as a s.t. c: x[2] = 3; # defined var with constant value 1/3 minimize zot: (x[1] - y)^2; 20161107 Fix a glitch (surprising error message) in 64-bit binaries with models having defined variables and at most one constraint or objective, but not one constraint and one objective. 20161110 Fix a parsing glitch (surprising error message) that could arise with a model or data command immediately following a compound command (at the outermost level) of the form for{i in ...} repeat { ... } while ... p[i] ... ; where the test in the while clause involves an entity subscripted by a dummy variable instantiated in the for{...}. 20161209 Fix a glitch with NaNs in min() and max() expressions. E.g., min(3,NaN) should be NaN, not 3. New command-line option -x sss instructs AMPL to run for at most about sss wall-clock seconds. It may run 4 or 5 seconds longer, particularly under MS Windows, to allow solvers to stop and clean up. On Unix-like systems, such as Linux, signals SIGHUP (when not being ignored, e.g., by running under nohup), SIGINT, SIGQUIT and SIGTERM are now passed to solvers, and SIGHUP (when not ignored) and SIGTERM cause the AMPL session to terminate. When the solver is a binary, the solver now sees all of SIGHUP, SIGINT, SIGQUIT and SIGTERM. When the solver is a shell script, results depend on what "trap" settings have been specified. New system parameters (automatically maintained, and not adjustable, e.g., in "let" commands): _session_time = wall-clock seconds consumed _session_maxtime = wall-clock session specified by "-x sss" When -x is not given on the AMPL command-line, _session_maxtime is reported to be 0. 20161220 Fix a bug with defined variables: nonlinear expressions all of whose variables were fixed by presolve were effectively ignored. Example: var x >= 0; var y = sqrt(x); s.t. c: x == 1; print y; # printed 0 rather than 1 20161231 Fix a rarely seen bug with defined variables fixed by presolve that led to the error message "bug: corrupted del_mblk arg". 20170111 Fix a bug with handling more than 64 output files from file redirections (">filename" or ">>filename"): a fault was possible in "solve" or "shell" commands on Unix-like systems, such as Linux. Adjust history processing on Unix-like systems so when the current input line is nonempty, control-D behave like "Delete", deleting the character under the cursor. Also arrange for the tab key to do filname-completion (when unambiguous, except on 32-bit Sparc Solaris). 20170126 Fix some bugs with tab completions, and extend ambiguous matches over the unambiguous part. 20170207 Fix a bug that could only bite a script running more than 4294967296 (i.e., 2^32) commands. 20170412 Under -Rw, disable history. Fix a glitch (surprising error message) in computing reduced costs involving an imported function with a string valued "if ... then ... else" argument. 20170531 Allow an optional comma between phrases in a set declaration (as already allowed in param, variable, objective, and constraint declarations -- none of which is described in the AMPL book). Allow commas to separate values in "read" commands. Fix a fault with complicated uses of piecewise-linear terms in iterated constraints. Fix some bugs (leading to invalid .nl files) with complicated uses of defined variables. 20170613 Fix another bug with defined variables; the bug resulted in an incorrect .nl file. Example: var x{i in 1..2} := i; var y1 = x[1] + x[2]^2; # nonlinear defined variable var y2 = 2*y1 + 1; # linear use of y1 minimize o: 1/y2; # nonlinear use of y2 20170614 Fix a fault that was possible under complicated conditions. 20170616 Honor "ordered" in set S{1..1} ordered; param p{S[1]}; data; param :S[1]: p := b 1 a 2 c 3; display p; # previously ignored the S[1]'s ordering and in set GEN ordered; param a {GEN}; table Gen IN /*...*/: [GEN] IN, a; read table Generators; display a; 20170621 In commands of the form for outerloop {...} { for {...} { ... commands foo; ... } } have a "break outerloop;" in file foo terminate the outer loop. 20170628 Fix a fault that arose under complicated conditions. Fix a nasty bug with defined variables that led to wrong objectives and constraints: when a defined variable had both a linear and a nonlinear part, wrong linear parts were sometimes transmitted to the .nl file under the default option linelim 1. Example: var x := 3; var y = 2*x + (x-1)^2; minimize foo: 3*x + y; # behaved as 5*x + y 20170630 Fix a botch in the changes of 20170621 that could cause a fault in a "commands" command. 20170706 Extend the change of 20170621 so in the context of for outerloop {...} { for {...} { ... commands foo; ... } } a "break outerloop;" in a file in a nest of "commands" commands in file foo will terminate the outer loop. 20170711 More fixes to "break" commands in nests of "commands" commands. 20170717 Fix a possible fault in "display p;" when p has an indexing set in which later set expressions depend on earlier ones. Example: set A; set B{A} ordered; param p{i in A, B[i]}; data; set A := 1 2; set B[1] := a b c; set B[2] := c a b; param p : a b c := 1 11 12 13 2 21 22 23; display p; # faulted 20170718 Fix a another possible fault in "display p;" when p has an indexing set involving three or more set expressions and some later set expressions depend on earlier ones. Whether the bug bit depended on the relative sizes of the component sets. Example: set A; set B{A} ordered; set C; param p{i in A, B[i], C} := Normal01(); data; set A := 1 2 3; set B[1] := a b c; set B[2] := c a b; set B[3] := b c a; set C := W X Y Z; display p; # faulted 20170724 Fix bugs with "break commands;" and "break all;" that could bite if the command terminated a loop with dummy variables. 20170731 Allow "and" as a synonym for "for all" and "or" as a synonym for "there exists". Thus the logical expressions there exists {i in I} b[i] and or {i in I} b[i] are treated alike, and the logical expressions for all all {i in I} b[i] and and {i in I} b[i] are treated alike. In "shell" and "solve" commands, ignore "> filename" and ">> filename" redirections when the file was specified in "option logfile" (which makes no sense). Previously a fault resulted. 20170902 Fix a bug with reading ".tab" tables with rows more than 2000 characters long. An occasional value might have been misread as 0. 20170914 Fix a bug with a subscripted defined variable subject to bound constraints when the indexing set, after being instantiated for one "solve", increased in size beyond the next power of 2 for a subsequent "solve". Add a Caution for "(numeric expression) & (numeric expression)", which converts both numeric values to strings and concatenates them. Since dummy variables may have string or numeric values, the new Caution is not given for "A & B" when A or B is a dummy variable. 20170924 Fix a fault that arose under unusual conditions. New debug options table_debug (default 0) and table_debug_template (default "%.dbtab") can cause tables in the format of .tab files to be read or written rather than or in addition to files specified for "read table" and "write table" commands. For this to work, option table_debug_template must contain one % character and otherwise only alphanumeric characters or plus, minus, underscore or period. When $table_debug and $table_debug_template indicate a .tab-format file, the file's name is obtained by substituting the declared table name for the % character in $table_debug_template; $table_debug is the sum of 1 ==> "read table" should read the .tab-format file and ignore the 2 bit if set in $table_debug 2 ==> "read table" should write the .tab-format file 4 ==> "write table" should also write the .tab-format file 8 ==> "write table" should not write the declared table While the builtin .tab file handlers only act on file names that end with ".tab", the .tab-format files controlled by $table_debug act on the constructed .tab-format file names. 20170925 In $table_debug_template just require one % and no white space. 20171002 Have printf format %q quote ".". Fix a glitch that sometimes caused "write table" to run presolve unnecessarily, which could cause a spurious "Solution determined by presolve" message to appear. Fix a bug in the handling of synonym _con by "write table". 20171028 Fix a bug with "ampl foo" when file foo ends with an incomplete data section (no final semicolon) whose last line is incomplete, i.e., does not end with a newline character. With some versions of MS Windows, the 64-bit ampl.exe faulted on such examples. 20171103 Fix a bug that, under complicated conditions involving "unfix", gave error message "... presolve finds no feasible solution possible." Example: var x >= 2 := 10; minimize f: x; fix x; s.t. c: x <= 5; display c.slack; unfix x; display c.slack; solve; # ... presolve finds no feasible solution possible. 20171107 Fix a glitch with "reset;" under -R: $PATH could have been corrupted. 20171111 Circumvent scurrilous behavior by Microsoft's ODBC software -- unexpectedly changing arithmetic details when x8087 instructions are involved. This was seen in a "write table" example. A "read table" likely would also have caused the trouble, which was seen in the round() function, but would also affect other binary <--> decimal conversions. 20171122 Fix a possible fault with "data foo" when file foo contains an ill-formed header. Extend option allow_NaN to apply to more contexts, e.g., computation of reduced costs when some constraints are removed by presolve. Example of using $allow_NaN in an interactive session: ampl: var x{1..2} >= 0; ampl: minimize zot: sum{i in 1..2} (x[i] - i)^2 + sqrt(x[1]); ampl: s.t. vex: x[1] + x[2] <= 1; ampl: s.t. c1: x[1] <= 0; ampl: solve; MINOS 5.51: optimal solution found. 1 iterations, objective 2 Nonlin evals: obj = 4, grad = 3. ampl: display x.rc; Error executing "display" command: Error differentiating zot: can't compute sqrt'(0). ampl: option allow_NaN 1; ampl: display x.rc; x.rc [*] := 1 NaN 2 0 ; 20171209 Fix unlikely trouble with reduced-cost computations when "option allow_NaN 1" is in effect and with exp(...) expressions that underflow to zero. Fix a bug with defined variables whose value is a linear expression plus a defined variable involving a nonconvex piecewise-linear term. The linear expression got lost. Example: var y >= -10; var z = 3*y; s.t. xb: -1 <= z <= 16; var pl = <<2,3,5,9,12;1,-1,1,-1,1,-1>> z; var c = 4*z + 3*pl; minimize pd: c; # the 4*z part got lost Fix another bug with nonconvex piecewise-linear terms illustrated by the example var x >= -1 <= 16; minimize pl: <<2,3,5,9,12;1,-1,1,-1,1,-1>> x; s.t. q: (x-6)^2 <= 200; which resulted in an invalid .nl file. Fix an infinite loop involving unusual use of defined variables. 20171212 Fix an obscure fault, hitherto not seen, involving piecewise linear terms. 20180115 Expand and clarify "ampl -v?" output. On Unix-like systems, extend history mechanism so escape-backspace acts like control-w, omitting the preceding word; control-a acts like the "Home" key, moving the cursor to the start of the line; control-e acts like the "End" key, moving the cursor to the end of the line; escape-b moves the cursor back one word; escape-f, like control-w, moves the cursor ahead one word; and escape-d omits the word ahead. On Unix-like systems, suppress history if at least one input file is given on the command line and no command-line input files are "-" (a single dash). This can be overridden by invoking "ampl -vi4 ...", which could matter if an "include -" phrase appears. 20180123 Fix a bug that could arise under complicated conditions with problems involving piecewise-linear terms or "var ... in set..." constructions. 20180125 Adjust history mechanism so it may deal better with unexpected escape sequences, and use a larger stacksize in hopes of avoiding a bug in Red Hat's glibc-2.17-196.el7_4.2.x86_64. 20180308 Under -b (block-mode), suppress superfluous command-identification lines that sometimes appeared. 20180316 Fix a bug with a sequence of solves involving fixing some but not all members of a set of variables, then unfixing all of the set of variables, then doing something causing the set of variables in question to be regenerated. The previously fixed variables were erroneously held fixed. The update of 20170706 seems to have revealed, but not caused, this bug. 20180416 Withdraw option funcwarn and the associated -f command-line option; now the former "option funcwarn 1" is always assumed. Fix a bug with the sequence solve; # error message about an unavailable function load some_library.dll; # to provide the function solve; The constraint or objective in question was not regenerated using the newly available function. Fix a bug with "load library" providing a decoding function: the "read" command was not properly handled after decoding started. Relevant documentation may only be provided on request. 20180417 "Invisible" tweak to yesterday's changes. 20180423 Fix a bug with a double inequality giving a bound on a defined variable and another constraint giving a stronger bound on the variable. The other bound on the variable was sometimes lost. 20180503 When single-stepping, try to recover from input errors rather than simply exiting. Fix a bug, introduced 20171122, with "param :setname: parname" followed by a : ... := table (called a value-table on p. 476 of the AMPL book). For example, set S dimen 2; param p{S}; data; param :S: p : X Y Z := A 1 2 3 B 4 . 6; display S, p; will now work again. 20180511 Fix a bug (possible fault) in processing utterly pointless input of the form "data; param p := .;". 20180519 Fix a rarely seen bug (e.g., losing some variable or constraint indices) that bit after an update, such as reordering a set, caused some entities to be reinstantiated. Compute tanh(x) and tanh'(x) for large x without complaint. 20190522 In data sections, accept "param :setname: parname (tr)" followed by a : ... := table (called a value-table on p. 476 of the AMPL book). For example, set S dimen 2; param p{S}; data; param :S: p (tr) : X Y Z := A 1 2 3 B 4 . 6; display S, p; will now work, with X, Y, and Z as first members of the tuples in S. 20180525 Compute tanh(x) and tanh'(x) for large |x| without complaint. 20180618 Fix a bug with string-valued imported functions that seems to have crept in with version 20070301: calls involving at least one numeric argument faulted. An example using amplfunc.dll compiled from http://ampl.com/netlib/ampl/solvers/funclink/funcadd.c: function kth symbolic; print kth(2,1,3,5); # faulted 20180624 Add variant "write 1;" of "write 0;" that does not emit the message "No files written: option outopt is 0." Have printf honor field widths given with %c format specifiers. For example, printf "X%4cY%-4cZ\n", 'a', 'b'; formerly printed "XaYbZ" and now prints "X aYb Z". 20180820 Fix a bug seen in scripts of the form repeat { solve; break; } print solve_exitcode; in which the "solve" commands gives a nonzero solve_exitcode. Modification to changes of 20160325: when a shell or solve command gives shell_exitcode > $shell_exitcode_max > 0 or solve_exitcode > $solve_exitcode_max > 0, respectively, show an option command that would have suppressed a "" message. The shell or solve command behaves as if a BREAK (control-C) had been received if $shell_exitcode > abs($shell_exitcode_max) or $solve_exitcode > abs($solve_exitcode_max), respectively. Another change: "exit code" is changed to "exit value" in some messages. When a shell or solve process ends, it returns an "exitcode value" that is assigned to option shell_exitcode or solve_exitcode, respectively, and on Unix-like systems (such as Linux and MacOSX) is computed as 256*(exit_value) + termination_code in which termination_code is zero if the shell or solve process ended by calling exit(exit_value) or returning exit_value and is nonzero if the exit() call could not be reached due to a segmentation fault (i.e., serious bug) or similar error. On MS Windows systems, the "exitcode value" is the exit_value. 20180822 Fix bugs with simplifying some logical constraints, leading to an error message about an "unexpected nonvariable type". 20180927 When an error message of the form z must be redeclared without y.val appears due to y being a defined variable, add text of the form because y is a defined variable and y.val is not allowed here. to the error message. 20181005 Fix a bug in combining linear expressions or with defined variables involving piecewise-linear terms. A fault was sometimes possible. 20181018 Fix a bug with limiting the number of file descriptors in use. When a file was closed so its file descriptor could be reused and later was reopened with a different file descriptor, a saved copy of the original descriptor was not updated. This caused a surprising read failure in the example that brought this bug to light. 20181022 Fix a rarely seen bug (fault) in saving variable values. Fix a bug with _ampl_time, _ampl_system_time, _ampl_user_time, _ampl_elapsed_time, and _session_time, which were not updated after their first evaluation. 20181026 Convert "param x random" to "var x random". Issue a Caution about this conversion unless "option randparam_warn 0;" has been specified. 20181102 Fix a bug introduced 20181022 in the logic that determines which entities need to be updated. Some were not updated when they should have been (in a complicated example). 20181114 Fix bugs (faults) with a "model" or "data" commands of the forms model (if t == 1 then 'a.mod' else 'b.mod') and model (if t == 1 then 'a.mod' else if t == 2 then 'b.mod') 20181123 Fix a seldom-seen bug in saving primal and dual variable values after the indexing set of the variable or constraint in question has decreased sufficiently in size. There was a fault in the illustrating example. Fix a fault with nonlinear use of a defined variable involving a piecewise-linear term. Example: var x in [-1,10]; var y = <<1,3,6;-1,1,-1,1>> x; minimize zot: exp(2*y); solve; # faulted 20181210 Fix a bug (fault or worse) with sequences of solves in which a variable or constraint indexing set starts with positive size, then is updated to have size 0, then is updated to have size larger than 1. 20181217 Fix a bug with iterated "let" in the following example: set S; var x{S}; let S := {}; /*empty: card(S) = 0*/ let{i in S} x[i] := i^2 + 1; let S := 1..4; /* now card(S) >= 2 */ let{i in S} x[i] := i^2 + 2; # corrupted memory 20190122 Fix a bug with data updates that cause some variables to disappear completely after being used in a "solve". A fault (or worse) was possible in a subsequent solve. 20190130 When presolve complains about an infeasible constraint with no variables, retain the constraint, so a second "write" or "solve" command will emit an infeasible problem when option infeas_clear has its default value 1, or when a .nl file is written by a command-line invocation of the form "ampl -ogfoo foo.mod". 20190207 New command sleep [expr]; causes the AMPL process to be suspended for expr seconds (default 1). Give a more detailed error message for attempts to use a function with out-args in a declaration. Fix a bug introduced in 20190130 that affected problems with piecewise-linear terms. Adjust the logic for "option randseed 0;" to avoid some trouble with repeated uses of this option: on some 64-bit systems, after a while "option randseed 0;" seemed not to affect the sequence of pseudo-random numbers generated. 20190220 Fix a fault with too few dummy variables in an indexing. Example: set S; set T = {t in {S,S}}; # faulted 20190223 Fix a bug (fault) with "reset options;" in a compound command, such as a loop. 20190418 In the contexts commands filename; data filename include filename model filename if filename starts with "<", the remainder of filename is taken to be a command and operands to be interpreted by the host operating system (e.g., by /bin/sh on Unix systems), and the standard output of the command is processed as the contents of filename. For example, model '< gzip -dc foo.mod.gz foo.dat.gz' and model '< unzip -p zap.zip zap.mod zap.dat' would decompress foo.mod.gz and foo.dat.gz or would extract zap.mod and zap.dat from zap.zip and take them as input. (For this to work, foo.dat or zap.dat would need to begin with "data;".) For "model", "data", and "commands" (but not "include") model < filename data < filename commands < filename; are treated as though filename began with "<", so, .e.g, model '< gzip -dc foo.mod.gz foo.dat.gz' and model < 'gzip -dc foo.mod.gz foo.dat.gz' are treated alike. 20190505 Redo update of 20190220 (to fix a fault with too few dummy variables in an indexing) so the arity of the indexing set is maintained. Adjust error processing so in interactive mode (reading from stdin when, e.g., -vi1 was specified on the command line or when stdin is a console window), execution of a command is not skipped during error recovery. 20190510 Fix a bug (possible fault) sometimes seen with logical expressions of the form lexpr1 ==> lexpr2 else lexpr3, where the lexprn are logical expressions. 20190529 Fix a bug with "display p;" involving declarations of the form set S; set T{S} ordered; param p{i in S, T[i]}; When the bug bit, it caused a surprising "display bug!" message. Fix a glitch with prompts. For example, after "include junk" typed at the "AMPL:" prompt, if file "junk" contained another "include", such as "data foo.dat", the "AMPL:" prompt was lost at the end of file "junk". 20190612 Fix a bug with variable.val: some expressions were not recomputed when the value of the variable changed. (Workaound: use variable rather than variable.val, which is also easier to read.) New option load_funcdcl determines whether imported functions are implicitly declared when their library is loaded or (new possibility) when an imported function makes another imported function available: 0 ==> no (default and the old behavior) 1 ==> yes, quietly 2 ==> yes, with a report of the declaration. Explicit declarations after implicit declarations are allowed. 20190616 Fix a bug with cleaning up after a "not defined" error in a nested-loop context. An infinite loop or surprising error message was possible. Example: set A dimen 2; data; set A := (1,a) (2,b); for{i in 1..2} for{j in {'a','b'}} print {k in U[j]} i,j,k; # U was not declared 20190617 Fix a bug (possible fault), introduced in version 20190418, with "option single_step 1; commands 'somefile';". Until further unforeseeable changes come along that require solvers to react to the AMPL version, the date that sometimes appears on line 1 of a .nl file is now fixed at 20190616. This should facilitate some kinds of testing, as it removes a gratuitous cause for differences in .nl files generated with different AMPL versions, at least those with versions >= 20190616. Forbid "load" commands from overriding existing declarations. 20190716 New builtin symbolic parameter _presolve_messages is assigned messages produced by presolve. New option show_presolve_messages (default 1) determines whether the presolve messages are printed. 20190814 Fix a bug with suffixes initialized in declarations: when some initial values were 0, it was possible for some subsequent initial values to be lost (converted to 0). Example: set A = 1..5; set C = {2,3,5}; var x {a in A} >= 0 integer suffix relax if a in C then 1 else 0; display x.relax; # x[5].relax was lost. Fix a glitch with Unicode (UTF8) characters affecting "show" commands and the output of "ampl -D" on input without commands. Some text was lost. Examples (requiring a window that understands UTF8): "ampl" on set pá; param cost {pá} > 0; show pá, cost; "ampl -D" on set pá; param cost {pá} > 0; data; set pá := dód cär; Fix a glitch introduced 20190716 with demo licenses: the error message for oversize problems was lost. 20190817 Complain when an input file ends in the middle of a comment. 20190819 Fix a glitch under MS Windows with block mode ("ampl -b ...") and option show_presolve_messages (at its default value). 20190824 Fix some seldom-seen bugs (faults or possible misbehavior), e.g., with slices in some command sequences or with bad data-section syntax. 20190830 Fix a bug with suffixes: after an indexed constraint goes from having some suffix values to none, due to a decrease in the indexing-set cardinality, a subsequent addition to the indexing set could lead to trouble. 20190911 Fix a rarely-seen glitch (wrong value computed) with a defined involving a piecewise-linear term applied to another defined variable involving a piecewise-linear term. Fix a fault that was possible with inappropriate outopt values, such as "write zfoo;". Invoking "ampl -o?" shows the allowed values. 20190919 Fix a potential compiler-dependent bug (not yet actually seen) on 64-bit systems with first(Set), last(Set), and card(Set). 20190927 Fix an obscure fault. 20191001 Fix bugs with "astatus" and "exitcode" suffixes on problems. Banish one reason for a rarely seen "OPDIV botch in e2v" message. 20191015 Fix memory leaks with "print", "printf", and "display". 20191031 Fix a bug that arose under complicated conditions involving the member() function. Arrange for "option cmdtrace 1;" to print command numbers with full precision, which only matters if many commands are executed. 20191108 Fix a glitch with blockmode ("ampl -b ...") introduced 20190819. Error messages from presolve had the wrong label. 20191116 Fix a bug with "drop" and "restore" that arose under complicated and rarely seen conditions. Memory corruption was possible. 20191223 Fix a bug with presolve on problems with piecewise-linear terms: an incorrect array reference could have an out-of-bounds subscript on problems of the "right" size. 20200110 Fix a fault with "option;" after adding a new option to another environment that is not in the current one. Add a Caution about ignoring a duplicate ":=" in a data section. The example that caused this addition led to a fault (now avoided). 20200123 Fix a seemingly invisible bug with "let _con[i] := ...". Fix a bug with \ in data sections. Example: param p symbolic; data; param p := "x\ \\ y"; # erroneous error message about missing " 20200501 Fix a bug with _nvars after a "delete variable" command. Fix an obscure display bug that gave error message "corrupted del_mblk arg". 20200810 Fix bugs in computing the value of a defined variable, which was sometimes wrong if the right-hand side involved a piecewise-linear term or a defined variable with a linear part. 20201019 Fix a bug (fault) with a set of variables indexed over the empty set. Have the delete command recover more memory from constraints and variables. 20201031 Fix a glitch (possible fault) in interactive mode connected with an error message for a wrong number of subscripts. Example: ampl: set S1 := 1..3; ampl: set S2 := 1..3; ampl: param a {S1,S2}; ampl: let a[1] := 1; # fault For entities with restricted domains (e.g., params with an "in" clause), check the restrictions at the end of "let". Example: set A := {'a', 'aa', 'aaa'}; param a symbolic in A; let a := 'bcd'; display a; # no longer reached 20201110 New option let_domain_check (default 0). The domain check in let, introduced 20201031, is now only done if $let_domain_check is nonzero. Someone showed us an example where the test greatly increased run time. 20201123 Fix a bug with some subscripted generic synonyms after some of the entities in question have been dropped/fixed. Wrong values were computed. Example: model diet.mod data diet.dat solve; drop diet['B1']; display _conname, _con, _con.body, _con.slack; display {i in 1.._ncons} (_conname[i], _con[i], _con[i].body, _con[i].slack); # _con[i].body and _con[i].slack were wrong for i > 1. Fix a bug with _ccon[something].astatus and _ccon[something].status, which gave "Bug: bad case 6 in f_OPSYNDOTSYM". 20210123 When too much memory is used, normally an error message such as Bug: Too much memory used -- 4276003800 bytes; couldn't get 32780 more. appears, but if a memory block needed to print how much memory was used cannot be allocated, simply print "Too much memory used." rather than silently quitting. 20210215 Fix a bug illustrated by: param p{i in 1..3} = if (i == 1) then 0 else p[i-1]+1; show p; display p; show p; # printed ":=" rather than "=" 20210220 Fix an error-message bug seen with interactive input. If file foo contains set A; var x{A}; s.t. c{i in A}: x[i] == 1; data; set A := a b c; then the interactive input ampl: include foo ampl: let{i in B} x[i] := 1; produced error messages B is not defined context: let{i in >>> B} <<< x[i] := 1; x is already defined context: let{i in B} >>> x[ <<< i] := 1; Today's change suppresses the erroneous error message about x. More generally, today's change only allows the first error message for a statement to appear. 20210226 Recant change of 20210220, as it suppressed syntax errors in interactive declarations. Fix a possible bug (e.g., fault) with recursive parameters. New command "reset check;" causes all "check" statements to be executed afresh by a "check" or "solve" or "solution" or "write" command. Normally a check statement is only executed at the first "check" or "solve" or "solution" or "write" command after the check statement's declaration or at a later "check" or "solve" or "solution" or "write" command if something the statement is testing has changed. 20210326 Fix a presolve bug with the default option var_bounds = 1 setting: if a constraint was dropped because it implied a stronger bound on an integer variable, the stronger bound was not conveyed to the solver. (Specifying "option var_bounds 2" was a work-around.) Fix a presolve bug with complementarity constraints. Removing a constraint after fixing all the variables in it should correctly resolve a complementarity condition involving the constraint. New builtin param _check_failures records how many check statements failed in the most recent relevant command ("check" or "solve" or "solution" or "write"). 20210330 Fix a bug with deleting constraints and variables that crept into version 20201019. The bug manifested itself at the second delete. 20210414 New option log_file_flush. If option log_file is given and $log_file_flush is 1, $log_file is flushed after each write to it. By default (option log_file_flush 0), $log_file is only flushed when the input file is stdin. 20210505 Fix bugs with data sections appearing after something causes an entity's declared default value to be used. Unless a "reset data" or "update data" command has given permission for the data section to provide a new value for the entity, there should be an error message. Example: var v default .1; display v; # cause default to be used data; var v := 0.08; # should (but did not) elicit an error message display v; # showed the old value, .1 # For sets, showed the new value. # With today's bug fix and with "reset data;" or "reset data p;" # or "update data;" or "update data p;" before "data;", # "display v;" would show the new value. 20210521 Fix a bug in the MS Windows binaries with "write ...; solve;". Binaries for other platforms (such as Linux and MacOS) are unaffected. 20210531 Fix a bug with complementarity constraints that could incorrectly lead to the message "presolve finds a nonsquare complementarity system", e.g., with econ.mod and econ.dat in the AMPL book. Change from 200 to 299 the value assigned to solve_result_num when presolve eliminates all variables and finds an infeasible constraint. Now whenever presolve finds the problem infeasible, solve_result_num should be set to 299. It is set to 99 when presolve solves a feasible problem. 20210613 Fix a couple of glitches in the "expand" and "solexpand" commands: incorrect printing (under complicated conditions) of comments "Eliminated by presolve" and "free row". 20210625 Fix a performance bug: after a "delete variable" command, some commands, such as "display", could cause presolve to run unnecessarily. 20210714 Fix another performance bug: assignments to suffixes, such as "relax", should not cause presolve to run in preparation for the assignment. When bailing out due to option eexit, dump presolve messages before stopping. Under MS Windows, allow blanks in option TMPDIR (a bad idea). Blanks in $TMPDIR were already allowed on other systems. 20210731 Fix an obscure bug: under complicated conditions, variable.val and constraint.val did not reflect recent changes. Example: var x; fix x := 3; display x.astatus; unfix x; let x := 5; display x.val, x; # printed x.val = 3 rather than 5 20210810 Fix a bug with split defined variables (where the linear part is sometimes handled separately) introduced in version 20190824. An example on which the bug caused a fault: var x := 1; var y = 2*x + x^2; c: x + y >= 1 complements sin(y) <= .5; Fix a bug that, under complicated conditions, affected some problems with complementarity constraints. 20210906 Fix a bug with suffixes: if the suffixed entity had not yet been instantiated, a fault was possible. Example: var x{i in 1..2} := i; var y{1..2}; s.t. c: sum{i in 1..2} x[i].val * y[i] >= 3; expand c; # faulted; workaround: "display x;" before "expand c;" Change to when an entity is instantiated: if the entity depends on suffix values, changes to those values no longer cause the entity to be instantiated anew. In the above example, adding let x[1] := 1.5; expand c; will give the same constraint as before -- the change to x[1] == x[1].val does not cause c to be reinstantiated. This change is partly to facilitate scripts that add cutting planes. It is also necessary for consistency with a forthcoming new implementation of AMPL (that supports functions in AMPL, among other things). AMPL has facilities for causing entities to be reinstantiated when declared prerequisites change. For example, changing the above example to use a named parameter for the coefficients of c will cause c to be reinstantiated when the parameter values change: param p{1..2}; redeclare s.t. c: sum{i in 1..2} p[i]*y[i] >= 3; let {i in 1..2} p[i] := x[i]; expand c; let p[1] := x[1] + 2; expand c; 20211222 After an error message "variable in ...", replace the whole expression in question by a constant (that can be seen with a "show" command). Previously the example var a := 12; set S := 4 .. a; display S; (if entered interactively) gave different results on different kinds of machines. Now "display S;" should give "set S := ; # empty" and "show S;" should give "set S = 4 .. 0;". 20220110 Fix a rarely seen bug with deducing bounds on variables, illustrated by var x >= 0; display x.ub; # correctly showed x.ub = Infinity var y = max{i in {0}} x;# logically the same as "var y = x;" display x.ub; # incorrectly showed x.ub = 0 Specifying "option presolve 0;" bypassed the bug. Fix a bug with handling certain errors. Example: ampl: param p {i in 1..10} := j; j is not defined context: param p {i in 1..10} := >>> j; <<< ampl: redeclare param p {i in 1..10} := i; Ignoring redeclaration of p: system parameters may not be redeclared. context: redeclare param p {i in 1..10} := >>> i; <<< ampl: display p; Segmentation fault (core dumped) Fix a bug with redeclare: set X; data; set X := 1; redeclare set X ordered; update data X; data; set X := 1 2; # faulted 20220119 Restore some error-handling behavior prior to 20220110. The changes of 20220110 sometimes resulted in several error messages per statement. Fix a parsing bug in error handling (infinite loop, introduced 20220110) seen with set A dimen 2; data; set A := (1,a) (2,b); for{i in 1..2} for{j in {'a','b'}} print {k in U[j]} i,j,k; # U was not declared 20220217 Fix a bug with "data filename;" in a compound statement. If a second such statement appeared in the compound statement with the same filename after the file was recreated after the first "data filename;", wrong data was read unless an explicit "close filename;" appeared before the file was recreated. Example (under Linux): set A; if (1 < 2) then { shell 'echo "set A := a b c;"' >foo; data foo # close foo; # uncomment to bypass bug display A; shell 'echo "set A := x y z;"' >foo; update data A; data foo display A; } For MS Windows, it is necessary to omit the double quotes in the shell commands and to add "close foo;" immediately after them. Then the example runs correctly. Fix an obscure fault with an input file without a final newline that is directly accessed by "include" at the command prompt. For example, if file foo contains just "param/" without a newline character, then "ampl foo" correctly said foo, line 1 (offset 5): unexpected end of file context: >>> / <<< but (in an interactive AMPL session) ampl: include foo faulted. 20220224 Fix an off-by-one error in line numbers on some error messages. For example, if "foo1" said "include foo2", foo2 said "model diet.mod" and the current directory did not have a file named diet.mod, then invoking ampl and typing "include foo1" gave foo2, line 1 (offset 6): Can't find file "diet.mod" context: model >>> diet.mod <<< include stack... -, line 0 includes foo1, line 0 includes foo2 rather than foo2, line 1 (offset 6): Can't find file "diet.mod" context: model >>> diet.mod <<< include stack... -, line 1 includes foo1, line 1 includes foo2 20220310 Fix more error-message bugs. If foo1 said "include foo2" and foo2 said "model diet.mod data diet.dat" and the current directory contained diet.mod but not diet.dat, then invoking ampl and typing "include foo1" gave a fault. After changing foo2 by moving "data diet.dat" to a second line, invoking ampl and typing "include foo1" gave foo2, line 2 (offset 159431244): Can't find file "diet.dat" context: data >>> diet.dat <<< ; include stack... -, line 1 includes foo1, line 1 includes foo2 with an erroneous offset. 20220323 Make option bad_subscripts apply in more cases. For example, set S; var x{S}; data; set S := a b; var x := b 2.1 a 1.1 c 3.1; display x; now gives Error executing "display" command: error processing var x: invalid subscript x['c'] discarded. x [*] := a 1.1 b 2.1 ; Inserting "option bad_subscripts 0;" before the display command suppresses the error message. The default $bad_subscripts is still 3. Fix a bug in displaying several items indexed over a cross-product of sets, some ordered, some not. A fault was sometimes possible. Example: set A ordered; set B; param x{A, B}; param y{A, A}; data; set A := 1; set B := 0; param x 1 0 1.1; param y 1 1 2.2; display x, y; # The 64-bit binaries faulted; 32-bit did not, # nor did "display y, x;". 20220505 Fix a bug in the "show" command, which did not print default (dual) values for constraints. Example: var x; s.t. cx default 1: x <= 4; show cx; Allow suffixes (on variables, constraints, objectives, problems) to appear in data sections. As usual for data sections, the name of a constraint, objective, or problem can be introduced with "param" or "var". For example, var x; minimize o: (x-3)^2; suffix foo; data; var o.foo 3.2; # or "var o := 3.2"; the ":=" is optional here. # or "param o.foo 3.2", etc. This is meant to make "snapshots" more efficient; there is no checking whether suffix values are replaced. When "option show_write_files 2" is specified and no variables remain used after presolve, print "# No .nl file written: no variables used." Quietly reduce absurdly large precisions in printf formats. For example, "%.410g" faulted. Now it works and "%.500g" is quietly reduced to "%.415g". 20220506 In data sections, when x is a variable, treat x.val as x when x does not yet have a current value, and similarly for c and c.dual when c is a constraint. In the little example var x; data; var x.dual := 3; change the error message from "dual is not a suffix" to "dual is not an assignable suffix." (When x is a defined variable, declared with "var x = expression;", x.dual is the value of the dual variable for the implied constraint "x = expression". Defined variables are substituted out of the problem, so the solver does not see them, but sometimes it is desirable to see dual variable values for them.) 20220525 Redo some binaries to incorporate the fix to printf of 20220505. Due to a botch, the fix was omitted from some binaries. If printf "%.500g\n", 1; does not fault, the fix made it into your AMPL binary. 20220621 Fix a glitch with console input: if the cursor is at the start of the current line (perhaps because of pressing the HOME key), the DELETE key did nothing. Increase the longest line that can be recalled correctly with history and the up and down arrow keys from 1000 to 4000 characters. 20220703 Fix a glitch introduced in version 20220505: in data sections, an unquoted "-" in the subscript of a subscripted set gave an error message. Example: set S; set T{S}; data; set S := 2022-06-26_03; set T[2022-06-26_03] := a b c; display T; Changing the third line to set T['2022-06-26_03'] := a b c; was a work-around. 20220730 Fix a bug with defined variables: if a defined variable used a problem variable nonlinearly, another defined variable used the first one linearly, no other use was made of the problem variable, and the second defined variable was used nonlinearly, then derivatives with respect to the problem variable were not computed. Example: set I = 1..3; var x{i in I} := i; var v1 = x[1] + x[2]^2; var v2 = v1 + x[3]^3; minimize obj: v2^2; Derivatives with respect to x[2] were not computed because of an error in the .nl file. 20220812 Fix a bug in the changes of 20220730 that caused a fault with some uses of defined variables -- only when the changes of 20220730 were relevant. 20220927 Fix error a couple of error messages caused by inappropriate uses of variables. Previously the (silly) example var x >= 0 integer; var y >= 0 integer; subj to con: x + y in {0,3,5}; solve; got the surprising error message presolve, constraint con: Logical constraint is always false. Infeasible constraints determined by presolve. Now the second line elicits the error message Cannot test whether a variable expression is in a set expression. context: subj to con: x + y in >>> {0,3,5}; <<< This example can be restated as var x >= 0 integer; var y >= 0 integer; var z in {0,3,5}; s.t. con: x + y == z; Previously the second line of the little example var x >= 0; subj to con: x in {0,3,5}; elicited the surprising error message continuous variable in tuple context: subj to con: x in >>> {0,3,5}; <<< Not it gets Cannot test whether a variable is in a set expression. context: subj to con: x in >>> {0,3,5}; <<< The example can be rewritten (without complaint) as var x in {0,3,5}; For a "var x in set_expr" declaration, AMPL quietly generates binary variables related by an SOS1 condition and a constraint defining x. 20221008 Fix a configuration bug in the 64-bit ARM binary that caused a wrong value for some things, such as Normal01(), to be computed. With today's 64-bit ARM binary, "ampl -vvq" will report version 20221008. 20221013 Fix a bug, introduced in version 20220505, in handling data sections. Example: set A := 1..2; param p{A}; param q{A}; data; param: p q := [*] 1 2.5 3 # erroneously complained about "." 2 4 5.6; display p, q; 20230317 Change the error message for display intersection{i in 1 .. 0} {i}; from "empty iterated intersection" to "empty iterator in iterated intersection". Make integer[1,10] equivalent to 1..10 instead of (erroneously) giving an error message. Fix a possible fault with "display _var;" on some systems. (If it does not fault, then there is not a problem here.) Fix a little off-by-one formatting bug in displaying some numeric values, as in display {a in 0..1, b in 0..1} (a, b, if a and b then 1 else 0); which gave : a b if a && b then 1 := 0 0 0 0 0 0 1 0 1 0 1 0 1 0 0 1 1 1 1 1 ; rather than : a b if a && b then 1 := 0 0 0 0 0 0 1 0 1 0 1 0 1 0 0 1 1 1 1 1 ; Show error context in blockmode (invocations of "ampl -b ..."). Fix some bugs with defined variables involving piecewise-linear terms. Examples: var x1 >= -100 := -100; var x2 >= -100 := -100; var x3 = 3*x1 - 2*x2 + 100; # should be 0 var x4 = 0.5*(3*x1 - 2*x2 - 100 + <<0;-1,1>>x3);# should be -100 var x5 = 3*x1 + 2*x2 - x4; # should be -400 var x6 = 0.5*(3*x1 + 2*x2 + x4 + <<0;-1,1>>x5); # should be -100, not -275 display x1, x2, x3, x4, x5, x6; reset; var x1 := 1; var x2 := 2; var x3 = 2*x1 - 3*x2; # should be -4 var x4 = x1 + 2*x2 + <<0;-1,1>> x3; # should be 9 var x5 = x1 + x2 + <<.5;1,2>>x4; # should be 20.5, not 10.5 var x6 = 2*x1 + 4*x2 + <<1;2,3>>x2; # should be 15 display x1, x2, x3, x4, x5, x6; After treating (by default) a constraint of the form variable = expression as acting like the variable were a defined variable, allow "let" to change the variable rather than having it be computed by the expression. In the following example, var x := 1 >= 0; var y; ydef: y = x + <<2;-1,2>> x; x was actually treated as the defined variable, leading to incorrect behavior after an assignment to x: option solver cplex; solve; CPLEX 20.1.0.0: optimal integer solution; objective 0 0 MIP simplex iterations 0 branch-and-bound nodes Objective = find a feasible point. ampl: display x, y; x = 2 y = 0 let x := 3; print x; #printed 2 rather than 3 Fix a bug in generated names for piecewise-linear terms and unions of intervals: when more than one set of constraints was generated, the same names were used. Now the generation number is appended, except for the first set. Also, change the character separating components of generated names from '+' to '@'. Example: set S := 1 .. 2; var x{S}; minimize zot: sum{i in S} <> x[i]; solexpand; produced minimize zot: -(zot+x[1]+s)[0] + (zot+x[1]+s)[1] - (zot+x[2]+s)[0] + (zot+x[2]+s)[1]; Now it produces minimize zot: -(zot@x[1]@s)[0] + (zot@x[1]@s)[1] - (zot@x[2]@s_2)[0] + (zot@x[2]@s_2)[1]; Fix a bug with "let". After, e.g., a "solve" or "solexpand", assignments of 0 to variables sometimes appeared to be ignored. Example: var x := 100; var z = 3*x + 2; solexpand; let x := 0; display x; # showed x = 100 rather than x = 0 Fix a rarely seen bug that some solvers say "integer infeasible" when a "hard" piecewise-linear term (neither convex nor concave in a context that would make the term "easy") might vanish. Example: var x1 >= -100 := -100; var x2 >= -100 := -100; var x3 = 3*x1 - 2*x2 + 100; var x4 = 0.5*(3*x1 - 2*x2 - 100 + <<0;-1,1>>x3); var x5 = 3*x1 + 2*x2 - x4 + 9; var x6; x6def: x6 = 0.5*(3*x1 + 2*x2 + x4 + <<0;-1,1>>x5); minimize o: x6; (Prior to the current bug fixes, an incorrect .nl file led to a wrong solution on this example.) Fix a bug (fault) in simplifying the right-operand of "or". Example: var x binary; var y binary; c: x = 0 or x = 0 and y = 0; fix0: y = 0; write 0; #faulted Fix bugs with supplying suffix values in a data section. Suffix values supplied this way might not have been transmitted to a .nl file or might have caused a fault. Fix a presolve bug that only mattered when there are two or more objectives and at least one piecewise-linear term. Change the message No variables used after elimination of defined variables. to No variables used. Fix a bug with the alldisjoint function, which should return 1 if its arguments are all distinct sets. (The sets may have some set members in common, but alldisjoint should return 1 if all the sets are distinct and 0 if two or more of the sets have all the same elements.) Fix a bug in evaluating defined variables in unusual circumstances, such as a "solve;" followed by an objective command for a previously declared objective, followed by a display command involving defined variables. Current variable values were sometimes not used in evaluating the defined variables. 20230407 Fix some bugs, such as a fault, introduced in version 20230317 that afflicted problems with complementarity constraints. Fix a glitch on the third line of the .nl header: when complementarity constraints were present and "option presolve 0;" was specified, nonlinear complementarities were counted as linear complementarities. Fix trouble with ord(a), next(a), and prev(a) when a runs over a subscripted set with subscripts involving dummy variables of a "for" loop. Example: set A; set B{A} ordered; data; set A := a b; set B[a] := c d; set B[b] := e f; for{a in A, b in B[a]} print a, b, ord(b); #faulted Change alldisjoint(...) to alldistinct(...), still true when ... are all different sets and false otherwise. Order is ignored, so alldistinct({'a','b'}, {'b','a'}) is false. New logical function alldisjoint(...) is true if ... are all disjoint sets and false otherwise. When linearizing piecewise-linear terms, honor values of declared variables. Previously introduced variables were presented to solvers with initial values of 0. Now they are presented with initial values that depend on the declared variables' current values. This may only matter to nonlinear solvers. 20230421 Fix a bug that prevented suffixes on logical constraints from being conveyed to the .nl file (except in the special case of no algebraic constraints, in which case the first suffix on a logical constraint was conveyed). Fix a bug introduced in 20230407 with certain uses of nested "for" loops. (It's more efficient to avoid "for" loops when iterated commands suffice.) Fix a bug with generation of integer intervals. Now the following works: set S := integer[0, 5]; display S; 20230426 Only print one "not within" error message (instead possibly of several). Example: set S; set A; set B{S} within A; data; set S := a b; set A := d e f g; set B[a] := h i; set B[b] := k l; display B; Disable some possibly inappropriate tests on entities assigned by "let". Examples that used to elicit error messages: param T; set A = 1..T; set B within A; let T := 6; let B := {i in A: i mod 2 = 0}; display A, B; let T := 5; let B := {i in A: i mod 3 = 0}; # elicited error msg display A, B; # now works correctly reset; # Begin another example param T; param p{1..2} != T; data; param p := 1 5 2 4; param T := 5; let p[1] := 2; # elicited error msg display p; # now works correctly 20230516 Fix bugs introduced 20230426 in some "let" statements for an entity declared with a side condition, such as a >= expression. 20230816 Fix a couple of bugs with debugging options -G and -O. (If you do not know what they do, you do not need them.) Fix a rarely seen memory-overwrite bug with multiple declared problems. 20230918 Fix a fault with reading suffix values in a data section. Example: var x{i in 1 .. 2}; minimize zot{j in 1 .. 3} : sum{i in 1 .. 2} (x[i] - (i + j))^2; data; param: x.sstatus_num := 1 2; var x := 1 2 2 3; Fix a fault simplifying logical expressions involving <==>. The bug only bit when the <==> had to be retained but its right-hand side could be simplified. In data sections, for a variable v and constraint c, let v.val denote the current (primal) value of v and let c.dual denote the current dual value of c. Hitherto v.val and v were treated alike, as were c.dual and c, in both cases giving initial values used for the first solve. This change is for the benefit of "snapshot". Make the $AMPLFUNC found at startup the $$AMPLFUNC value, so it survives a "reset option;" command. New debugging option -w causes AMPL to give return code 0 when exiting after producing various error messages. Fix a glitch (fault) illustrated by var x{i in 1..2} integer := i; c: x[1] < x[2]; # logical constraint print c; #faulted Note that if c is a logical constraint, then c.val and c should be treated alike, but changing "print c;" to "print c.val;" in the above example avoided the fault. When 'e' appears in ($solver & '_auxfiles'), so that a .env file containing the current environment is written by the solve command, remove the .env file at the end of the solve, like other temporary files, such as the .nl file. When 'e' appears in $auxfiles, have the write command write a .env file. 20231012 Fix a glitch in reading values in solution (.sol) files for suffixes on logical constraints. Fix a bug with a new problem after solving a problem with a "var ... in set_expr" declaration when the new problem does not involve that variable. Constraints and variables used to implement the "in set_expr" phrase were erroneously included in the new problem. 20231129 Fix a small performance bug in logical constraints that allowed, e.g., "|| 0" to creep into the .nl file in unusual cases. An example is complicated. Fix a bug (seen in a complicated example) in the ordering of _sconnames; entities should be in declare order. When expanding logical constraints by expand or solexpand commands, force the constraint body onto a new line. Do not show the solver logical constraints that are always true. New builtin suffix .lno for constraints is positive for logical constraints seen by solvers, 0 for algebraic constraints seen by solvers, whereas .sno is now 0 for logical constraints seen by solvers and positive for algebraic constraints seen by solvers. Thus if c is declared a logical constraint, c.lno + c.sno > 0 if the constraint is seen by solvers. If presolve turns c into an algebraic constraint, then c.sno > 0. If c.lno > 0, then _slogcon[c.lno] is the logical constraint presented to solvers, and if c.sno > 0, then _scon[c.sno] is the algebraic constraint presented to solvers. Fix a solexpand bug with logical constraints converted by presolve to algebraic constraints. Example: var x{1..3} >= 0; c1: x[1] = .5; c2: x[1] > .4 ==> x[1]^2 + x[2]^2 >= 1; minimize zot: 2*x[2] + 3*x[3]; solexpand; # said subject to c2:x[2]^2 + 0.75; # rather than #subject to c2: # x[2]^2 >= 0.75; In the example just given, "display c2.lno, c2.sno;" now shows c2.lno = 0 c2.sno = 1 New builtin suffix .domain for variables to indicate the kind of values a variable can take: 0 ==> floating-point (double precision) 1 ==> integer 2 ==> in a discrete set 3 ==> in a union of intervals The default .domain is 0, but a variable's declaration can specify the other possibilities. Specifying a variable's domain to be a single interval gives .domain 0 and has the same effect as giving explicit lower and upper bounds. Integer variables can have nonintegral values, e.g., due to a "let" command or a "solve" command, and can have a nonzero .relax suffix to indicate that solvers should view them as continuous. Binary variables are integer variables with lower bound 0 and upper bound 1. Additional feature: it seems we neglected to describe the count operator in the AMPL book. It has one of the forms count(list_of_expressions) count {indexing} (list_of_expressions) In the second form, dummies from the indexing can be used in the list_of_expressions. Both forms return the number of expressions that are true (if logical) or nonzero (if arithmetic). 20240105 Fix a bug, introduced in version 20231129, in writing .row files for problems containing logical constraints. (Most problems do not involve such constraints, and .row files are not written by default). For example, in the (silly) script set I = 1..9; var x{I}; function myfunc; var t{i in 1..3} = x[i]^2 + 1 + sum{j in 8..9} (i+j)*x[j]; var u{i in 1..2} = x[7+i]^2 + 2 + sinh(x[1] + 2*t[2] + 6*x[6]); maximize zip: if t[2] >= 0 then -t[2]^3 else -t[2]^2; minimize zap: sin(t[1]) + cos(2*t[2]) + 4*x[4] + 5*x[5] + x[6]^2 + x[7]^2; minimize zot: cosh(<<1,2;3,4,5>>x[6]); c1: t[2] + sin(t[3]) <= 4; c2: x[5] + cos(x[6]) >= 3; c3: sum{i in 3..7} i*x[i] = 1; c4: 4.3 <= x[5] + myfunc(t[2], x[3]*x[6], 'some string') <= 15.5; b45{i in 4..5}: x[i] >= i; b1: x[1] <= 3.5; b2: -1 <= x[2] <= 2; b3{i in 8..9}: 0 <= x[i] <= 0.1*i; lc{i in 1..2}: x[6] + x[7] >= 2.5 ==> (x[5] + x[6]^2)^2 + u[i] <= 35; suffix zork; let{i in 2.._nvars} _var[i].zork := i; option nl_comments 1, auxfiles rc; write gsilly; the resulting silly.row file contained "=u[1]" and "=t[1]" rather than "lc[1]" and "lc[2]". Adjust the third integer in lines that start with V in g format .nl files to accord with the description in https://ampl.github.io/nlwrite.pdf. Whether that integer is 0 or nonzero is all that really matters and is not affected by this change. Fix a bug (fault) with "display _slogcon;" added to the end of the above example. Fix a fault with ord applied to some dummy variables. Example: set I ordered := 1..10; for {i in I, j in I: ord(i) < ord(j)} display i, j; 20240110 New builtin function tableindexarity('foo') analogous to indexarity('foo'), but relevant to table declations, since tables have their own name space. Thus the (silly) script set A = 1..3; param p{A}; table test IN: [A], p; table test2{a in A} IN: [index] p; display indexarity("test"), indexarity("test2"), tableindexarity("test"), tableindexarity("test2"); param test{A,A,A}; display indexarity("test"); gives output indexarity('test') = -1 indexarity('test2') = -1 tableindexarity('test') = 0 tableindexarity('test2') = 1 indexarity('test') = 3 20240116 Fix a glitch in the new tableindexarity function that depended on whether a table's name was an even or odd member of _TABLES, illustrated by set A = 1..3; param p{A}; table test1{A} IN: [index] p; table test2{A,A} IN: [index] p; display tableindexarity('test1'); # gave 0 rather than 1 display tableindexarity('test2'); New builtin function environindexarity, analogous to indexarity and tableindexarity, but for environments. Example: environ foo; set A; environ goo{A}; environ zoo{A,A,A}; display environindexarity('foo'), environindexarity('goo'), environindexarity('zoo'); # environindexarity('foo') = 0 # environindexarity('goo') = 1 # environindexarity('zoo') = 3 20240208 Fix a glitch in converting 0x1.fffffffffffffp-1023 to binary. (This is very unlikely to affect any real problems.) The little test script param p; data; param p := 0x1.fffffffffffffp-1023; display p, 0x1.fffffffffffffp-1023; gave p = 0 0 = 0 rather than p = 2.22507e-308 2.2250738585072014e-308 = 2.22507e-308 Fix a bug, introduced in 20240105, with problems having logical constraints and an objective. After a "solve" or "solution" command, a further "write" without changes omitted "G" segments in the .nl file. 20240224 Fix a glitch introduced in version 20231129 seen under "option presolve 0" (often a bad idea) in problems having complementarity constraints. 20240308 Add the possibility of $auxfiles containing 'd', which is treated like 'c' to cause write commands to emit a .col file that has the names of defined variables used in the .nl file to appear at the end of the .col file in the order b, c, o, c1, o1, where b = number of defined variables used in both a constraint and an objective; c = number of defined variables used in two or more constraints but no objectives; o = number of defined variables used in two or more objectives but no constraints; c1 = number of defined variables just used in one constraint; o1 = number of defined variables just used in one objective. Fix a glitch with printing of expressions involving "prev". Note that prev(S,n) is rendered as next(S,-n). Example: set S := {1, 2, 3, 4, 5} ordered; var x{S} >= 0, integer; s.t. con{i in S: i > 2}: x[prev(i,2)] <= x[i]; param p := prev(3, S, 2); show con; show p; should print subject to con{i in S: i > 2} : x[next(i, S, -2)] <= x[i]; param p = next(3, S, -2); rather than subject to con{i in S: i > 2} : x[next(i, S, 2)] <= x[i]; param p = next(3, S, 2); 20240313 Fix a fault with write commands when $auxfiles contains "d" and a defined variable is split into linear and nonlinear parts. In the .col file, nonlinear parts are now indicated by a ".nl" suffix. In the following silly example var x; var y; var z = 4*y + y^4 - 6.38; var w = 8*x + abs(y) + 18.635; minimize O: 5*x + 8*z + 15; s.t. C: -200 <= 205*x -17*x^3 - 3.8*w <= 403.2869; solexpand; option auxfiles rd; write gfoo1; the nonlinear part of z is y^4 and is denoted by z.nl, and the nonlinear part of w is abs(y) and is denoted by w.nl in the solexpand output and in foo1.col. 20240519 Fix a bug with logical constraints eliminated by presolve that led to an error message of the form "Bug: presolve has k = ...". Fix a fault seen in a complicated example involving two solves. Fix a bug introduced in version 20231017 in handling complementarity constraints in which an inequality could be expressed in a variable declaration. Example (from dempe.mod by Sven Leyffer): var x; var z; var w; minimize f: (x - 3.5)^2 + (z + 4)^2; c1: z - 3 + 2*z*w = 0; c2: 0 <= w complements x - z^2 >= 0; Solvers sometimes erroneously found this problem to be infeasible. The bug depended only bit under some conditions; for example, removing f or c1 in the above example made the bug disappear. New bit 128 in pl_linearize to permit defined variables involving piecewise-linear expressions to be substituted out, which may lead to slight numerical differences. Example: var X >= 0; var Y >= 0; C: Y = 2*X + X^2; D: Y^2 <= 10; minimize zot: X + Y + exp(X + Y) + <<1,2,3;-1,1,-1,1>> Y; option solver knitro; solve; Knitro 12.4.0: Locally optimal or satisfactory solution. objective 0.9999999997; integrality gap 3.16e-10 1 nodes; 2 subproblem solves # ... option pl_linearize 129; solve; Knitro 12.4.0: Locally optimal or satisfactory solution. objective 1; integrality gap 0 1 nodes; 2 subproblem solves Fix bug introduced in version 20240105 in conveying suffix values of logical constraints to .nl files. 20240524 Fix a bug introduced 20240519 with logical constraints. Example: # Adapted from: # www.g12.csse.unimelb.edu.au/minizinc/downloads/examples-latest/oss.mzn param duration {1..3}; var Start {1..3} integer >= 0, <= 2809; var Makespan integer >= 0, <= 2809; minimize Objective: Makespan; NoMachineConflicts {m1 in 1..3, m2 in m1+1..3}: Start[m1] + duration[m1] <= Start[m2] or Start[m2] + duration[m2] <= Start[m1]; MakespanDefn {m in 1..3}: Start[m] + duration[m] <= Makespan; data; param duration:= 1 121 2 333 3 343 ; solexpand; # Incorrectly showed NoMachineConflicts[1,2] thrice and omitted # other subscripts of NoMachineConflicts. 20240531 Fix a recently introduced bug (fault) seen with the little example var x; var y; maximize o: x + 2*y; c: x <= 3 complements y <= 4; 20240619 Fix a recently introduced bug in models with logical and algebraic constraints. Wrong Jacobian details appeared in the .nl file. Fix possible trouble in an error message about possible suffix names. 20240805 Fix obscure bugs a couple of obscure printing bugs: "print 6.9999999999999996e-161;" gave 7e161 (a one-bit error with the usual IEEE 64-bit double arithmetic) and "print .000491831912544516;" gave 0.0004918319125445159 (which with the same IEEE double arithmetic rounds to the same value but is longer than necessary). Fix a bug with logical constraints seen under complicated conditions. A simple example seems hard to find. Fix a bug seen in the script var x; data; var x.val := 4.2; display x; display x, x.val; # printed zeros Similarly, fix obscure bugs seen in the scripts var x; suffix glop; data; var x.init := 4.2; var x.val := 3.7; var x.glop := 5.1; display x; display x, x.val, x.init, x.glop; # printed 0 rather than 3.7 for x and x.val update data; data; var x.init := 7.4; var x.val := 13.3; var x.glop := 6.1; display x, x.val, x.init, x.glop; # printed 0 for x, x.val, and x.init and var x{1..3}; suffix glop; data; var x.init := 1 4.2 2 4.3 3 4.4; var x.val := 1 3.7 2 3.8 3 3.9; var x.glop := 1 5.1 2 5.2 3 5.3; display x; display x, x.val, x.init, x.glop; # printed wrong values for x and x.val update data; data; var x.init := 1 7.4 2 7.5 3 7.6; var x.val := 1 13.3 2 13.4 3 13.5; var x.glop := 1 6.1 2 6.2 3 6.3; display x, x.val, x.init, x.glop; # printed wrong values update data; data; var : x.init x.val x.glop := 1 1.1 2.1 3.1 2 1.2 2.2 3.2 3 1.3 2.3 3.3 ; display x, x.val, x.init, x.glop; # printed wrong values Fix bugs computing dual values for constraints eliminated by presolve when logical constraints are present. For example, set I = 1..4; var x{I} >= 0; c1: x[1] + x[2] >= 2 ==> x[3] + x[4] <= 3; c2: x[2] = 1; minimize zot: sum{i in I} i*x[i]; write 0; # causes presolve to run display c2; # said 0 rather than 2 Another example (involving another bug) is set I = 1..5; var x{I} >= 0; c1{j in 1..2}: x[j] + x[3] >= 2 ==> x[3] + x[4] <= 3; c2{j in 2..3}: x[j] = 1; minimize zot: sum{i in I} i*x[i]; write 0; display c2; # gave c2[3] = 0 rather than 3 Fix a bug presolving complementarity constraints: if the complementarity condition is resolved because one constraint holds with equality, the other must still be enforced. Example: set I = 1..5; var x{I}; c: x[1] + 2*x[2] >= 3 complements x[1] + x[3] + 2*x[4] + x[5]^2 <= 7; f{i in 1..4}: x[i] = 1; solexpand c; # showed a free row rather than a constraint Fix a fault seen in set I = 1..6; var x{I}; c: x[1] + 2*x[2] >= 3 complements x[1] + x[3] + 2*x[4] + x[5]^2 <= 7; f{i in 1..4}: x[i] = 1; l: x[5]^2 >= 1 ==> x[6]^2 <= 2; solexpand c, l; # faulted 20240825 Fix a bug with suffix values in two data sections separated by a command involving the first suffix. Example: model diet.mod data diet.dat suffix sos2; data; var Buy.sos2 := BEEF 1 CHK 1 FISH 2 HAM 1 MCH 2 SPG 2 TUR 2; display Buy.sos2; suffix zork; data; var Buy.zork := BEEF 1 CHK 1.5 FISH 2.3; display Buy.zork; # displayed all zeros Fix a bug with specifying v.val in a data section for a scalar (i.e., unsubscripted) variable v. Example: var x := 4.3; data; param x.val = 3.2; display x.val; # showed 0 instead of 3.2 Fix a glitch introduced 20240805 in handling declared initial variables under some conditions. Example: var x{i in 1..2} := i + 1.5; var y{i in 1..2} = x[i]^2 + 1; # bug not shown with "display x;" let x[1] := 4; display x, y; # Showed x[2] = 0 rather than 3.5. # To see the bug, need to display both x and y. 20240826 Fix a bug introduced 20230317 with defined variables used in piecewise-linear terms. Fix a bug that sometimes bit with general piecewise-linear terms (i.e., those not convex in a context where convexity can be exploited and are not concave in a context where concavity can be exploited) introduced 20240519. An example afflicted by both bugs: var x >= 0 <= 10; var z = <<1;-1,1>> x; c: <<1.1, 2.2; -1, .2, 1>> x + z*z >= 5.41; 20241028 On machines with arithmetic conforming to IEEE standard 754 (most modern machines) option allow_NaN 2; now permits handling Infinity and NaN (Not a Number) as specified in the IEEE standard, whereas option allow_NaN 1; turns Infinity into NaN and the default option allow_NaN 0; causes computations that would involve NaNs and Infinities to give error messages. "option allow_NaN 1" has long been available; "option allow_NaN 2" is new. For example, if file krf defines and uses a rational function # See W. Kahan (1987), "Presubstitution, and Continued Fractions" # http://www.arithmazium.org/classroom/lib/Kahan_Presubstitution_and_Continued_Fractions.pdf var x; var f = 4 - 3/(x - 2 - 1/(x - 7 + 10/(x - 2 - 2/(x-3)))); for{i in 1..4} { let x := i; display x, f; } then by default "ampl krf" says x = 1 Error at _cmdno 3 executing "display" command (file f/krf, line 7, offset 255): error computing defined variable f: can't compute 10/0 but the input "option allow_NaN 2; include krf" gives x = 1 f = 7 x = 2 f = 4 x = 3 f = 1.6 x = 4 f = 2.5 and "option allow_NaN 1; include krf" gives x = 1 f = NaN x = 2 f = NaN x = 3 f = NaN x = 4 f = NaN If file zork contains option allow_NaN 2; display exp(3000); option allow_NaN 1; display exp(3000); option allow_NaN 0; display exp(3000); then "ampl zork" gives exp(3000) = Infinity exp(3000) = NaN Error at _cmdno 6 executing "display" command (file zork, line 6, offset 98): can't evaluate exp(3000): Numerical result out of range Fix some bugs with a defined variable involving a piecewise-linear term applied to another defined variable that involves another piecewise-linear term, possibly indirectly. Examples: var x1 >= -100 := -100; var x2 >= -100 := -100; var x3 = 3*x1 - 2*x2 + 100; var x4; x4def: x4 = 0.5*(3*x1 - 2*x2 - 100 + <<0;-1,1>>x3); var x5 = 3*x1 + 2*x2 - x4 + 9; minimize o: 0.5*(3*x1 + 2*x2 + x4 + <<0;-1,1>>x5); option solver cplex; solve; CPLEX 20.1.0.0: Constraint _scon[1] is not convex quadratic since it is an equality constraint. With the bug fixed, "solve;" says CPLEX 20.1.0.0: optimal integer solution; objective -104.5 Another, similar but more troublesome example is the above with x4 as a defined variable, i.e., var x1 >= -100 := -100; var x2 >= -100 := -100; var x3 = 3*x1 - 2*x2 + 100; var x4 = 0.5*(3*x1 - 2*x2 - 100 + <<0;-1,1>>x3); var x5 = 3*x1 + 2*x2 - x4 + 9; minimize o: 0.5*(3*x1 + 2*x2 + x4 + <<0;-1,1>>x5); After the fixes, "option solver cplex; solve;" says CPLEX 20.1.0.0: optimal integer solution; objective -104.5 2 MIP simplex iterations 0 branch-and-bound nodes Fix a bug (corrupted del_mblk arg) with ord(i) when i is a dummy variable. The bug appeared in a complicated example; some simple examples work correctly. Allow "from", "to" and some more obscure values to appear as dummy variables. Example: set A dimen 2; var x{(to,from) in A: to != from}; previously elicited a syntax error message. Now it is accepted.