LAPACK 3.3.1
Linear Algebra PACKage
|
00001 REAL FUNCTION CLA_SYRPVGRW( UPLO, N, INFO, A, LDA, AF, LDAF, IPIV, 00002 $ WORK ) 00003 * 00004 * -- LAPACK routine (version 3.2.2) -- 00005 * -- Contributed by James Demmel, Deaglan Halligan, Yozo Hida and -- 00006 * -- Jason Riedy of Univ. of California Berkeley. -- 00007 * -- June 2010 -- 00008 * 00009 * -- LAPACK is a software package provided by Univ. of Tennessee, -- 00010 * -- Univ. of California Berkeley and NAG Ltd. -- 00011 * 00012 IMPLICIT NONE 00013 * .. 00014 * .. Scalar Arguments .. 00015 CHARACTER*1 UPLO 00016 INTEGER N, INFO, LDA, LDAF 00017 * .. 00018 * .. Array Arguments .. 00019 COMPLEX A( LDA, * ), AF( LDAF, * ) 00020 REAL WORK( * ) 00021 INTEGER IPIV( * ) 00022 * .. 00023 * 00024 * Purpose 00025 * ======= 00026 * 00027 * CLA_SYRPVGRW computes the reciprocal pivot growth factor 00028 * norm(A)/norm(U). The "max absolute element" norm is used. If this is 00029 * much less than 1, the stability of the LU factorization of the 00030 * (equilibrated) matrix A could be poor. This also means that the 00031 * solution X, estimated condition numbers, and error bounds could be 00032 * unreliable. 00033 * 00034 * Arguments 00035 * ========= 00036 * 00037 * UPLO (input) CHARACTER*1 00038 * = 'U': Upper triangle of A is stored; 00039 * = 'L': Lower triangle of A is stored. 00040 * 00041 * N (input) INTEGER 00042 * The number of linear equations, i.e., the order of the 00043 * matrix A. N >= 0. 00044 * 00045 * INFO (input) INTEGER 00046 * The value of INFO returned from CSYTRF, .i.e., the pivot in 00047 * column INFO is exactly 0. 00048 * 00049 * NCOLS (input) INTEGER 00050 * The number of columns of the matrix A. NCOLS >= 0. 00051 * 00052 * A (input) COMPLEX array, dimension (LDA,N) 00053 * On entry, the N-by-N matrix A. 00054 * 00055 * LDA (input) INTEGER 00056 * The leading dimension of the array A. LDA >= max(1,N). 00057 * 00058 * AF (input) COMPLEX array, dimension (LDAF,N) 00059 * The block diagonal matrix D and the multipliers used to 00060 * obtain the factor U or L as computed by CSYTRF. 00061 * 00062 * LDAF (input) INTEGER 00063 * The leading dimension of the array AF. LDAF >= max(1,N). 00064 * 00065 * IPIV (input) INTEGER array, dimension (N) 00066 * Details of the interchanges and the block structure of D 00067 * as determined by CSYTRF. 00068 * 00069 * WORK (input) COMPLEX array, dimension (2*N) 00070 * 00071 * ===================================================================== 00072 * 00073 * .. Local Scalars .. 00074 INTEGER NCOLS, I, J, K, KP 00075 REAL AMAX, UMAX, RPVGRW, TMP 00076 LOGICAL UPPER 00077 COMPLEX ZDUM 00078 * .. 00079 * .. Intrinsic Functions .. 00080 INTRINSIC ABS, REAL, AIMAG, MAX, MIN 00081 * .. 00082 * .. External Subroutines .. 00083 EXTERNAL LSAME, CLASET 00084 LOGICAL LSAME 00085 * .. 00086 * .. Statement Functions .. 00087 REAL CABS1 00088 * .. 00089 * .. Statement Function Definitions .. 00090 CABS1( ZDUM ) = ABS( REAL ( ZDUM ) ) + ABS( AIMAG ( ZDUM ) ) 00091 * .. 00092 * .. Executable Statements .. 00093 * 00094 UPPER = LSAME( 'Upper', UPLO ) 00095 IF ( INFO.EQ.0 ) THEN 00096 IF ( UPPER ) THEN 00097 NCOLS = 1 00098 ELSE 00099 NCOLS = N 00100 END IF 00101 ELSE 00102 NCOLS = INFO 00103 END IF 00104 00105 RPVGRW = 1.0 00106 DO I = 1, 2*N 00107 WORK( I ) = 0.0 00108 END DO 00109 * 00110 * Find the max magnitude entry of each column of A. Compute the max 00111 * for all N columns so we can apply the pivot permutation while 00112 * looping below. Assume a full factorization is the common case. 00113 * 00114 IF ( UPPER ) THEN 00115 DO J = 1, N 00116 DO I = 1, J 00117 WORK( N+I ) = MAX( CABS1( A( I, J ) ), WORK( N+I ) ) 00118 WORK( N+J ) = MAX( CABS1( A( I, J ) ), WORK( N+J ) ) 00119 END DO 00120 END DO 00121 ELSE 00122 DO J = 1, N 00123 DO I = J, N 00124 WORK( N+I ) = MAX( CABS1( A( I, J ) ), WORK( N+I ) ) 00125 WORK( N+J ) = MAX( CABS1( A( I, J ) ), WORK( N+J ) ) 00126 END DO 00127 END DO 00128 END IF 00129 * 00130 * Now find the max magnitude entry of each column of U or L. Also 00131 * permute the magnitudes of A above so they're in the same order as 00132 * the factor. 00133 * 00134 * The iteration orders and permutations were copied from csytrs. 00135 * Calls to SSWAP would be severe overkill. 00136 * 00137 IF ( UPPER ) THEN 00138 K = N 00139 DO WHILE ( K .LT. NCOLS .AND. K.GT.0 ) 00140 IF ( IPIV( K ).GT.0 ) THEN 00141 ! 1x1 pivot 00142 KP = IPIV( K ) 00143 IF ( KP .NE. K ) THEN 00144 TMP = WORK( N+K ) 00145 WORK( N+K ) = WORK( N+KP ) 00146 WORK( N+KP ) = TMP 00147 END IF 00148 DO I = 1, K 00149 WORK( K ) = MAX( CABS1( AF( I, K ) ), WORK( K ) ) 00150 END DO 00151 K = K - 1 00152 ELSE 00153 ! 2x2 pivot 00154 KP = -IPIV( K ) 00155 TMP = WORK( N+K-1 ) 00156 WORK( N+K-1 ) = WORK( N+KP ) 00157 WORK( N+KP ) = TMP 00158 DO I = 1, K-1 00159 WORK( K ) = MAX( CABS1( AF( I, K ) ), WORK( K ) ) 00160 WORK( K-1 ) = 00161 $ MAX( CABS1( AF( I, K-1 ) ), WORK( K-1 ) ) 00162 END DO 00163 WORK( K ) = MAX( CABS1( AF( K, K ) ), WORK( K ) ) 00164 K = K - 2 00165 END IF 00166 END DO 00167 K = NCOLS 00168 DO WHILE ( K .LE. N ) 00169 IF ( IPIV( K ).GT.0 ) THEN 00170 KP = IPIV( K ) 00171 IF ( KP .NE. K ) THEN 00172 TMP = WORK( N+K ) 00173 WORK( N+K ) = WORK( N+KP ) 00174 WORK( N+KP ) = TMP 00175 END IF 00176 K = K + 1 00177 ELSE 00178 KP = -IPIV( K ) 00179 TMP = WORK( N+K ) 00180 WORK( N+K ) = WORK( N+KP ) 00181 WORK( N+KP ) = TMP 00182 K = K + 2 00183 END IF 00184 END DO 00185 ELSE 00186 K = 1 00187 DO WHILE ( K .LE. NCOLS ) 00188 IF ( IPIV( K ).GT.0 ) THEN 00189 ! 1x1 pivot 00190 KP = IPIV( K ) 00191 IF ( KP .NE. K ) THEN 00192 TMP = WORK( N+K ) 00193 WORK( N+K ) = WORK( N+KP ) 00194 WORK( N+KP ) = TMP 00195 END IF 00196 DO I = K, N 00197 WORK( K ) = MAX( CABS1( AF( I, K ) ), WORK( K ) ) 00198 END DO 00199 K = K + 1 00200 ELSE 00201 ! 2x2 pivot 00202 KP = -IPIV( K ) 00203 TMP = WORK( N+K+1 ) 00204 WORK( N+K+1 ) = WORK( N+KP ) 00205 WORK( N+KP ) = TMP 00206 DO I = K+1, N 00207 WORK( K ) = MAX( CABS1( AF( I, K ) ), WORK( K ) ) 00208 WORK( K+1 ) = 00209 $ MAX( CABS1( AF( I, K+1 ) ), WORK( K+1 ) ) 00210 END DO 00211 WORK( K ) = MAX( CABS1( AF( K, K ) ), WORK( K ) ) 00212 K = K + 2 00213 END IF 00214 END DO 00215 K = NCOLS 00216 DO WHILE ( K .GE. 1 ) 00217 IF ( IPIV( K ).GT.0 ) THEN 00218 KP = IPIV( K ) 00219 IF ( KP .NE. K ) THEN 00220 TMP = WORK( N+K ) 00221 WORK( N+K ) = WORK( N+KP ) 00222 WORK( N+KP ) = TMP 00223 END IF 00224 K = K - 1 00225 ELSE 00226 KP = -IPIV( K ) 00227 TMP = WORK( N+K ) 00228 WORK( N+K ) = WORK( N+KP ) 00229 WORK( N+KP ) = TMP 00230 K = K - 2 00231 ENDIF 00232 END DO 00233 END IF 00234 * 00235 * Compute the *inverse* of the max element growth factor. Dividing 00236 * by zero would imply the largest entry of the factor's column is 00237 * zero. Than can happen when either the column of A is zero or 00238 * massive pivots made the factor underflow to zero. Neither counts 00239 * as growth in itself, so simply ignore terms with zero 00240 * denominators. 00241 * 00242 IF ( UPPER ) THEN 00243 DO I = NCOLS, N 00244 UMAX = WORK( I ) 00245 AMAX = WORK( N+I ) 00246 IF ( UMAX /= 0.0 ) THEN 00247 RPVGRW = MIN( AMAX / UMAX, RPVGRW ) 00248 END IF 00249 END DO 00250 ELSE 00251 DO I = 1, NCOLS 00252 UMAX = WORK( I ) 00253 AMAX = WORK( N+I ) 00254 IF ( UMAX /= 0.0 ) THEN 00255 RPVGRW = MIN( AMAX / UMAX, RPVGRW ) 00256 END IF 00257 END DO 00258 END IF 00259 00260 CLA_SYRPVGRW = RPVGRW 00261 END