Annotation of rpl/lapack/lapack/dgelsx.f, revision 1.14

1.9       bertrand    1: *> \brief <b> DGELSX solves overdetermined or underdetermined systems for GE matrices</b>
                      2: *
                      3: *  =========== DOCUMENTATION ===========
                      4: *
                      5: * Online html documentation available at 
                      6: *            http://www.netlib.org/lapack/explore-html/ 
                      7: *
                      8: *> \htmlonly
                      9: *> Download DGELSX + dependencies 
                     10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dgelsx.f"> 
                     11: *> [TGZ]</a> 
                     12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dgelsx.f"> 
                     13: *> [ZIP]</a> 
                     14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dgelsx.f"> 
                     15: *> [TXT]</a>
                     16: *> \endhtmlonly 
                     17: *
                     18: *  Definition:
                     19: *  ===========
                     20: *
                     21: *       SUBROUTINE DGELSX( M, N, NRHS, A, LDA, B, LDB, JPVT, RCOND, RANK,
                     22: *                          WORK, INFO )
                     23: * 
                     24: *       .. Scalar Arguments ..
                     25: *       INTEGER            INFO, LDA, LDB, M, N, NRHS, RANK
                     26: *       DOUBLE PRECISION   RCOND
                     27: *       ..
                     28: *       .. Array Arguments ..
                     29: *       INTEGER            JPVT( * )
                     30: *       DOUBLE PRECISION   A( LDA, * ), B( LDB, * ), WORK( * )
                     31: *       ..
                     32: *  
                     33: *
                     34: *> \par Purpose:
                     35: *  =============
                     36: *>
                     37: *> \verbatim
                     38: *>
                     39: *> This routine is deprecated and has been replaced by routine DGELSY.
                     40: *>
                     41: *> DGELSX computes the minimum-norm solution to a real linear least
                     42: *> squares problem:
                     43: *>     minimize || A * X - B ||
                     44: *> using a complete orthogonal factorization of A.  A is an M-by-N
                     45: *> matrix which may be rank-deficient.
                     46: *>
                     47: *> Several right hand side vectors b and solution vectors x can be
                     48: *> handled in a single call; they are stored as the columns of the
                     49: *> M-by-NRHS right hand side matrix B and the N-by-NRHS solution
                     50: *> matrix X.
                     51: *>
                     52: *> The routine first computes a QR factorization with column pivoting:
                     53: *>     A * P = Q * [ R11 R12 ]
                     54: *>                 [  0  R22 ]
                     55: *> with R11 defined as the largest leading submatrix whose estimated
                     56: *> condition number is less than 1/RCOND.  The order of R11, RANK,
                     57: *> is the effective rank of A.
                     58: *>
                     59: *> Then, R22 is considered to be negligible, and R12 is annihilated
                     60: *> by orthogonal transformations from the right, arriving at the
                     61: *> complete orthogonal factorization:
                     62: *>    A * P = Q * [ T11 0 ] * Z
                     63: *>                [  0  0 ]
                     64: *> The minimum-norm solution is then
                     65: *>    X = P * Z**T [ inv(T11)*Q1**T*B ]
                     66: *>                 [        0         ]
                     67: *> where Q1 consists of the first RANK columns of Q.
                     68: *> \endverbatim
                     69: *
                     70: *  Arguments:
                     71: *  ==========
                     72: *
                     73: *> \param[in] M
                     74: *> \verbatim
                     75: *>          M is INTEGER
                     76: *>          The number of rows of the matrix A.  M >= 0.
                     77: *> \endverbatim
                     78: *>
                     79: *> \param[in] N
                     80: *> \verbatim
                     81: *>          N is INTEGER
                     82: *>          The number of columns of the matrix A.  N >= 0.
                     83: *> \endverbatim
                     84: *>
                     85: *> \param[in] NRHS
                     86: *> \verbatim
                     87: *>          NRHS is INTEGER
                     88: *>          The number of right hand sides, i.e., the number of
                     89: *>          columns of matrices B and X. NRHS >= 0.
                     90: *> \endverbatim
                     91: *>
                     92: *> \param[in,out] A
                     93: *> \verbatim
                     94: *>          A is DOUBLE PRECISION array, dimension (LDA,N)
                     95: *>          On entry, the M-by-N matrix A.
                     96: *>          On exit, A has been overwritten by details of its
                     97: *>          complete orthogonal factorization.
                     98: *> \endverbatim
                     99: *>
                    100: *> \param[in] LDA
                    101: *> \verbatim
                    102: *>          LDA is INTEGER
                    103: *>          The leading dimension of the array A.  LDA >= max(1,M).
                    104: *> \endverbatim
                    105: *>
                    106: *> \param[in,out] B
                    107: *> \verbatim
                    108: *>          B is DOUBLE PRECISION array, dimension (LDB,NRHS)
                    109: *>          On entry, the M-by-NRHS right hand side matrix B.
                    110: *>          On exit, the N-by-NRHS solution matrix X.
                    111: *>          If m >= n and RANK = n, the residual sum-of-squares for
                    112: *>          the solution in the i-th column is given by the sum of
                    113: *>          squares of elements N+1:M in that column.
                    114: *> \endverbatim
                    115: *>
                    116: *> \param[in] LDB
                    117: *> \verbatim
                    118: *>          LDB is INTEGER
                    119: *>          The leading dimension of the array B. LDB >= max(1,M,N).
                    120: *> \endverbatim
                    121: *>
                    122: *> \param[in,out] JPVT
                    123: *> \verbatim
                    124: *>          JPVT is INTEGER array, dimension (N)
                    125: *>          On entry, if JPVT(i) .ne. 0, the i-th column of A is an
                    126: *>          initial column, otherwise it is a free column.  Before
                    127: *>          the QR factorization of A, all initial columns are
                    128: *>          permuted to the leading positions; only the remaining
                    129: *>          free columns are moved as a result of column pivoting
                    130: *>          during the factorization.
                    131: *>          On exit, if JPVT(i) = k, then the i-th column of A*P
                    132: *>          was the k-th column of A.
                    133: *> \endverbatim
                    134: *>
                    135: *> \param[in] RCOND
                    136: *> \verbatim
                    137: *>          RCOND is DOUBLE PRECISION
                    138: *>          RCOND is used to determine the effective rank of A, which
                    139: *>          is defined as the order of the largest leading triangular
                    140: *>          submatrix R11 in the QR factorization with pivoting of A,
                    141: *>          whose estimated condition number < 1/RCOND.
                    142: *> \endverbatim
                    143: *>
                    144: *> \param[out] RANK
                    145: *> \verbatim
                    146: *>          RANK is INTEGER
                    147: *>          The effective rank of A, i.e., the order of the submatrix
                    148: *>          R11.  This is the same as the order of the submatrix T11
                    149: *>          in the complete orthogonal factorization of A.
                    150: *> \endverbatim
                    151: *>
                    152: *> \param[out] WORK
                    153: *> \verbatim
                    154: *>          WORK is DOUBLE PRECISION array, dimension
                    155: *>                      (max( min(M,N)+3*N, 2*min(M,N)+NRHS )),
                    156: *> \endverbatim
                    157: *>
                    158: *> \param[out] INFO
                    159: *> \verbatim
                    160: *>          INFO is INTEGER
                    161: *>          = 0:  successful exit
                    162: *>          < 0:  if INFO = -i, the i-th argument had an illegal value
                    163: *> \endverbatim
                    164: *
                    165: *  Authors:
                    166: *  ========
                    167: *
                    168: *> \author Univ. of Tennessee 
                    169: *> \author Univ. of California Berkeley 
                    170: *> \author Univ. of Colorado Denver 
                    171: *> \author NAG Ltd. 
                    172: *
                    173: *> \date November 2011
                    174: *
                    175: *> \ingroup doubleGEsolve
                    176: *
                    177: *  =====================================================================
1.1       bertrand  178:       SUBROUTINE DGELSX( M, N, NRHS, A, LDA, B, LDB, JPVT, RCOND, RANK,
                    179:      $                   WORK, INFO )
                    180: *
1.9       bertrand  181: *  -- LAPACK driver routine (version 3.4.0) --
1.1       bertrand  182: *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
                    183: *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
1.9       bertrand  184: *     November 2011
1.1       bertrand  185: *
                    186: *     .. Scalar Arguments ..
                    187:       INTEGER            INFO, LDA, LDB, M, N, NRHS, RANK
                    188:       DOUBLE PRECISION   RCOND
                    189: *     ..
                    190: *     .. Array Arguments ..
                    191:       INTEGER            JPVT( * )
                    192:       DOUBLE PRECISION   A( LDA, * ), B( LDB, * ), WORK( * )
                    193: *     ..
                    194: *
                    195: *  =====================================================================
                    196: *
                    197: *     .. Parameters ..
                    198:       INTEGER            IMAX, IMIN
                    199:       PARAMETER          ( IMAX = 1, IMIN = 2 )
                    200:       DOUBLE PRECISION   ZERO, ONE, DONE, NTDONE
                    201:       PARAMETER          ( ZERO = 0.0D0, ONE = 1.0D0, DONE = ZERO,
                    202:      $                   NTDONE = ONE )
                    203: *     ..
                    204: *     .. Local Scalars ..
                    205:       INTEGER            I, IASCL, IBSCL, ISMAX, ISMIN, J, K, MN
                    206:       DOUBLE PRECISION   ANRM, BIGNUM, BNRM, C1, C2, S1, S2, SMAX,
                    207:      $                   SMAXPR, SMIN, SMINPR, SMLNUM, T1, T2
                    208: *     ..
                    209: *     .. External Functions ..
                    210:       DOUBLE PRECISION   DLAMCH, DLANGE
                    211:       EXTERNAL           DLAMCH, DLANGE
                    212: *     ..
                    213: *     .. External Subroutines ..
                    214:       EXTERNAL           DGEQPF, DLAIC1, DLASCL, DLASET, DLATZM, DORM2R,
                    215:      $                   DTRSM, DTZRQF, XERBLA
                    216: *     ..
                    217: *     .. Intrinsic Functions ..
                    218:       INTRINSIC          ABS, MAX, MIN
                    219: *     ..
                    220: *     .. Executable Statements ..
                    221: *
                    222:       MN = MIN( M, N )
                    223:       ISMIN = MN + 1
                    224:       ISMAX = 2*MN + 1
                    225: *
                    226: *     Test the input arguments.
                    227: *
                    228:       INFO = 0
                    229:       IF( M.LT.0 ) THEN
                    230:          INFO = -1
                    231:       ELSE IF( N.LT.0 ) THEN
                    232:          INFO = -2
                    233:       ELSE IF( NRHS.LT.0 ) THEN
                    234:          INFO = -3
                    235:       ELSE IF( LDA.LT.MAX( 1, M ) ) THEN
                    236:          INFO = -5
                    237:       ELSE IF( LDB.LT.MAX( 1, M, N ) ) THEN
                    238:          INFO = -7
                    239:       END IF
                    240: *
                    241:       IF( INFO.NE.0 ) THEN
                    242:          CALL XERBLA( 'DGELSX', -INFO )
                    243:          RETURN
                    244:       END IF
                    245: *
                    246: *     Quick return if possible
                    247: *
                    248:       IF( MIN( M, N, NRHS ).EQ.0 ) THEN
                    249:          RANK = 0
                    250:          RETURN
                    251:       END IF
                    252: *
                    253: *     Get machine parameters
                    254: *
                    255:       SMLNUM = DLAMCH( 'S' ) / DLAMCH( 'P' )
                    256:       BIGNUM = ONE / SMLNUM
                    257:       CALL DLABAD( SMLNUM, BIGNUM )
                    258: *
                    259: *     Scale A, B if max elements outside range [SMLNUM,BIGNUM]
                    260: *
                    261:       ANRM = DLANGE( 'M', M, N, A, LDA, WORK )
                    262:       IASCL = 0
                    263:       IF( ANRM.GT.ZERO .AND. ANRM.LT.SMLNUM ) THEN
                    264: *
                    265: *        Scale matrix norm up to SMLNUM
                    266: *
                    267:          CALL DLASCL( 'G', 0, 0, ANRM, SMLNUM, M, N, A, LDA, INFO )
                    268:          IASCL = 1
                    269:       ELSE IF( ANRM.GT.BIGNUM ) THEN
                    270: *
                    271: *        Scale matrix norm down to BIGNUM
                    272: *
                    273:          CALL DLASCL( 'G', 0, 0, ANRM, BIGNUM, M, N, A, LDA, INFO )
                    274:          IASCL = 2
                    275:       ELSE IF( ANRM.EQ.ZERO ) THEN
                    276: *
                    277: *        Matrix all zero. Return zero solution.
                    278: *
                    279:          CALL DLASET( 'F', MAX( M, N ), NRHS, ZERO, ZERO, B, LDB )
                    280:          RANK = 0
                    281:          GO TO 100
                    282:       END IF
                    283: *
                    284:       BNRM = DLANGE( 'M', M, NRHS, B, LDB, WORK )
                    285:       IBSCL = 0
                    286:       IF( BNRM.GT.ZERO .AND. BNRM.LT.SMLNUM ) THEN
                    287: *
                    288: *        Scale matrix norm up to SMLNUM
                    289: *
                    290:          CALL DLASCL( 'G', 0, 0, BNRM, SMLNUM, M, NRHS, B, LDB, INFO )
                    291:          IBSCL = 1
                    292:       ELSE IF( BNRM.GT.BIGNUM ) THEN
                    293: *
                    294: *        Scale matrix norm down to BIGNUM
                    295: *
                    296:          CALL DLASCL( 'G', 0, 0, BNRM, BIGNUM, M, NRHS, B, LDB, INFO )
                    297:          IBSCL = 2
                    298:       END IF
                    299: *
                    300: *     Compute QR factorization with column pivoting of A:
                    301: *        A * P = Q * R
                    302: *
                    303:       CALL DGEQPF( M, N, A, LDA, JPVT, WORK( 1 ), WORK( MN+1 ), INFO )
                    304: *
                    305: *     workspace 3*N. Details of Householder rotations stored
                    306: *     in WORK(1:MN).
                    307: *
                    308: *     Determine RANK using incremental condition estimation
                    309: *
                    310:       WORK( ISMIN ) = ONE
                    311:       WORK( ISMAX ) = ONE
                    312:       SMAX = ABS( A( 1, 1 ) )
                    313:       SMIN = SMAX
                    314:       IF( ABS( A( 1, 1 ) ).EQ.ZERO ) THEN
                    315:          RANK = 0
                    316:          CALL DLASET( 'F', MAX( M, N ), NRHS, ZERO, ZERO, B, LDB )
                    317:          GO TO 100
                    318:       ELSE
                    319:          RANK = 1
                    320:       END IF
                    321: *
                    322:    10 CONTINUE
                    323:       IF( RANK.LT.MN ) THEN
                    324:          I = RANK + 1
                    325:          CALL DLAIC1( IMIN, RANK, WORK( ISMIN ), SMIN, A( 1, I ),
                    326:      $                A( I, I ), SMINPR, S1, C1 )
                    327:          CALL DLAIC1( IMAX, RANK, WORK( ISMAX ), SMAX, A( 1, I ),
                    328:      $                A( I, I ), SMAXPR, S2, C2 )
                    329: *
                    330:          IF( SMAXPR*RCOND.LE.SMINPR ) THEN
                    331:             DO 20 I = 1, RANK
                    332:                WORK( ISMIN+I-1 ) = S1*WORK( ISMIN+I-1 )
                    333:                WORK( ISMAX+I-1 ) = S2*WORK( ISMAX+I-1 )
                    334:    20       CONTINUE
                    335:             WORK( ISMIN+RANK ) = C1
                    336:             WORK( ISMAX+RANK ) = C2
                    337:             SMIN = SMINPR
                    338:             SMAX = SMAXPR
                    339:             RANK = RANK + 1
                    340:             GO TO 10
                    341:          END IF
                    342:       END IF
                    343: *
                    344: *     Logically partition R = [ R11 R12 ]
                    345: *                             [  0  R22 ]
                    346: *     where R11 = R(1:RANK,1:RANK)
                    347: *
                    348: *     [R11,R12] = [ T11, 0 ] * Y
                    349: *
                    350:       IF( RANK.LT.N )
                    351:      $   CALL DTZRQF( RANK, N, A, LDA, WORK( MN+1 ), INFO )
                    352: *
                    353: *     Details of Householder rotations stored in WORK(MN+1:2*MN)
                    354: *
1.8       bertrand  355: *     B(1:M,1:NRHS) := Q**T * B(1:M,1:NRHS)
1.1       bertrand  356: *
                    357:       CALL DORM2R( 'Left', 'Transpose', M, NRHS, MN, A, LDA, WORK( 1 ),
                    358:      $             B, LDB, WORK( 2*MN+1 ), INFO )
                    359: *
                    360: *     workspace NRHS
                    361: *
                    362: *     B(1:RANK,1:NRHS) := inv(T11) * B(1:RANK,1:NRHS)
                    363: *
                    364:       CALL DTRSM( 'Left', 'Upper', 'No transpose', 'Non-unit', RANK,
                    365:      $            NRHS, ONE, A, LDA, B, LDB )
                    366: *
                    367:       DO 40 I = RANK + 1, N
                    368:          DO 30 J = 1, NRHS
                    369:             B( I, J ) = ZERO
                    370:    30    CONTINUE
                    371:    40 CONTINUE
                    372: *
1.8       bertrand  373: *     B(1:N,1:NRHS) := Y**T * B(1:N,1:NRHS)
1.1       bertrand  374: *
                    375:       IF( RANK.LT.N ) THEN
                    376:          DO 50 I = 1, RANK
                    377:             CALL DLATZM( 'Left', N-RANK+1, NRHS, A( I, RANK+1 ), LDA,
                    378:      $                   WORK( MN+I ), B( I, 1 ), B( RANK+1, 1 ), LDB,
                    379:      $                   WORK( 2*MN+1 ) )
                    380:    50    CONTINUE
                    381:       END IF
                    382: *
                    383: *     workspace NRHS
                    384: *
                    385: *     B(1:N,1:NRHS) := P * B(1:N,1:NRHS)
                    386: *
                    387:       DO 90 J = 1, NRHS
                    388:          DO 60 I = 1, N
                    389:             WORK( 2*MN+I ) = NTDONE
                    390:    60    CONTINUE
                    391:          DO 80 I = 1, N
                    392:             IF( WORK( 2*MN+I ).EQ.NTDONE ) THEN
                    393:                IF( JPVT( I ).NE.I ) THEN
                    394:                   K = I
                    395:                   T1 = B( K, J )
                    396:                   T2 = B( JPVT( K ), J )
                    397:    70             CONTINUE
                    398:                   B( JPVT( K ), J ) = T1
                    399:                   WORK( 2*MN+K ) = DONE
                    400:                   T1 = T2
                    401:                   K = JPVT( K )
                    402:                   T2 = B( JPVT( K ), J )
                    403:                   IF( JPVT( K ).NE.I )
                    404:      $               GO TO 70
                    405:                   B( I, J ) = T1
                    406:                   WORK( 2*MN+K ) = DONE
                    407:                END IF
                    408:             END IF
                    409:    80    CONTINUE
                    410:    90 CONTINUE
                    411: *
                    412: *     Undo scaling
                    413: *
                    414:       IF( IASCL.EQ.1 ) THEN
                    415:          CALL DLASCL( 'G', 0, 0, ANRM, SMLNUM, N, NRHS, B, LDB, INFO )
                    416:          CALL DLASCL( 'U', 0, 0, SMLNUM, ANRM, RANK, RANK, A, LDA,
                    417:      $                INFO )
                    418:       ELSE IF( IASCL.EQ.2 ) THEN
                    419:          CALL DLASCL( 'G', 0, 0, ANRM, BIGNUM, N, NRHS, B, LDB, INFO )
                    420:          CALL DLASCL( 'U', 0, 0, BIGNUM, ANRM, RANK, RANK, A, LDA,
                    421:      $                INFO )
                    422:       END IF
                    423:       IF( IBSCL.EQ.1 ) THEN
                    424:          CALL DLASCL( 'G', 0, 0, SMLNUM, BNRM, N, NRHS, B, LDB, INFO )
                    425:       ELSE IF( IBSCL.EQ.2 ) THEN
                    426:          CALL DLASCL( 'G', 0, 0, BIGNUM, BNRM, N, NRHS, B, LDB, INFO )
                    427:       END IF
                    428: *
                    429:   100 CONTINUE
                    430: *
                    431:       RETURN
                    432: *
                    433: *     End of DGELSX
                    434: *
                    435:       END

CVSweb interface <joel.bertrand@systella.fr>