This directory contains a collection of test problems for nonlinear programming expressed as AMPL models (the *.mod files). Many are for a fixed problem size and contain a data section. Some are pure model files describing families of problems, with default values that specify a small instance; such models are noted with "V" in the classification table below. Comments in the *.mod files give the problem sources (original sources or test-problem collections). Many of the problems were selected and coded in AMPL by Elena Bobrovnikova (at Bell Labs in the summer of 1996), with some subsequent editing by David M. Gay; these problems contain the comment # Original AMPL coding by Elena Bobrovnikova (summer 1996 at Bell Labs). Each problem is classified as in the CUTE collection of test problems, except that problems with a V code for the number of variables or constraints have the default value for that number after the V. The classification is a string of the form OCSD-KI-n-m in which the letters are replaced by appropriate codes, as follows: O = objective code: C = Constant L = Linear N = No objective Q = Quadratic S = Sum of squares O = Other (none of the above) C = constraint type: B = Bounds on variables L = Linear N = linear Network Q = Quadratic U = Unconstrained X = just fiXed variables O = Other (none of the above) S = smoothness: R = Twice continuously differentiable I = Other D = degree of available derivatives: always 2 here; 0, 1, or 2 in general K = kind of the problem: A = Academic M = Modeling R = Real application I = internal variables: Y = yes, problems has useful internal variables N = no useful internal variables n = number of variables; V = varies (followed by default) m = number of constraints; V = varies (followed by default) Problem Classification blend.mod LOR2-MN-24-38 branin.mod OBR2-AN-2-4 camel1.mod OBR2-AN-2-4 chemeq.mod OLR2-MY-38-50 chi.mod OBR2-AN-2-4 gold.mod OBR2-AN-2-4 gridneta.mod OLR2-MY-V60-V96 griewank.mod OBR2-AN-2-4 hs105.mod OBR2-RN-8-16 hs106.mod LQR2-MN-8-22 hs109.mod OOR2-MY-9-26 hs111.mod OOR2-MN-10-23 hs112.mod OLR2-MY-10-13 hs114.mod QOR2-MY-10-31 hs116.mod LQR2-MN-13-41 hs15.mod OQR2-MN-2-3 hs23.mod QQR2-MN-2-9 hs35.mod QLR2-MN-3-4 hs44.mod QLR2-MN-4-10 hs5.mod OBR2-MN-2-4 hs54.mod OLR2-MN-6-13 hs6.mod QQR2-RN-2-1 hs62.mod OLR2-MY-3-7 hs64.mod OOR2-MN-3-4 hs8.mod CQR2-RN-2-2 hs87.mod LOR2-RN-11-24 kowalik.mod OBR2-AN-4-8 levy3.mod OBR2-AN-2-4 ljcluster.mod OUR2-AY-V20-0 osborne1.mod OBR2-MN-5-10 powell.mod OBR2-AY-4-4 price.mod OBR2-AN-2-4 rosenbr.mod OUR2-AN-2-0 s324.mod QQR2-AN-2-3 s383.mod OLR2-AN-14-29 schwefel.mod OBR2-AN-5-10 shekel.mod OUR2-AN-4-8 steenbre.mod ONR2-MY-540-666 tre.mod OBR2-AN-2-4 weapon.mod OLR2-MN-100-147 File minos.out shows results of using MINOS 5.5 to solve the problems on an SGI computer, with primal and dual values displayed by the AMPL commands display _varname, _var; display _conname, _con;