LAPACK  3.10.0
LAPACK: Linear Algebra PACKage
ddrvls.f
Go to the documentation of this file.
1 *> \brief \b DDRVLS
2 *
3 * =========== DOCUMENTATION ===========
4 *
5 * Online html documentation available at
6 * http://www.netlib.org/lapack/explore-html/
7 *
8 * Definition:
9 * ===========
10 *
11 * SUBROUTINE DDRVLS( DOTYPE, NM, MVAL, NN, NVAL, NNS, NSVAL, NNB,
12 * NBVAL, NXVAL, THRESH, TSTERR, A, COPYA, B,
13 * COPYB, C, S, COPYS, NOUT )
14 *
15 * .. Scalar Arguments ..
16 * LOGICAL TSTERR
17 * INTEGER NM, NN, NNB, NNS, NOUT
18 * DOUBLE PRECISION THRESH
19 * ..
20 * .. Array Arguments ..
21 * LOGICAL DOTYPE( * )
22 * INTEGER MVAL( * ), NBVAL( * ), NSVAL( * ),
23 * $ NVAL( * ), NXVAL( * )
24 * DOUBLE PRECISION A( * ), B( * ), C( * ), COPYA( * ), COPYB( * ),
25 * $ COPYS( * ), S( * )
26 * ..
27 *
28 *
29 *> \par Purpose:
30 * =============
31 *>
32 *> \verbatim
33 *>
34 *> DDRVLS tests the least squares driver routines DGELS, DGETSLS, DGELSS, DGELSY,
35 *> and DGELSD.
36 *> \endverbatim
37 *
38 * Arguments:
39 * ==========
40 *
41 *> \param[in] DOTYPE
42 *> \verbatim
43 *> DOTYPE is LOGICAL array, dimension (NTYPES)
44 *> The matrix types to be used for testing. Matrices of type j
45 *> (for 1 <= j <= NTYPES) are used for testing if DOTYPE(j) =
46 *> .TRUE.; if DOTYPE(j) = .FALSE., then type j is not used.
47 *> The matrix of type j is generated as follows:
48 *> j=1: A = U*D*V where U and V are random orthogonal matrices
49 *> and D has random entries (> 0.1) taken from a uniform
50 *> distribution (0,1). A is full rank.
51 *> j=2: The same of 1, but A is scaled up.
52 *> j=3: The same of 1, but A is scaled down.
53 *> j=4: A = U*D*V where U and V are random orthogonal matrices
54 *> and D has 3*min(M,N)/4 random entries (> 0.1) taken
55 *> from a uniform distribution (0,1) and the remaining
56 *> entries set to 0. A is rank-deficient.
57 *> j=5: The same of 4, but A is scaled up.
58 *> j=6: The same of 5, but A is scaled down.
59 *> \endverbatim
60 *>
61 *> \param[in] NM
62 *> \verbatim
63 *> NM is INTEGER
64 *> The number of values of M contained in the vector MVAL.
65 *> \endverbatim
66 *>
67 *> \param[in] MVAL
68 *> \verbatim
69 *> MVAL is INTEGER array, dimension (NM)
70 *> The values of the matrix row dimension M.
71 *> \endverbatim
72 *>
73 *> \param[in] NN
74 *> \verbatim
75 *> NN is INTEGER
76 *> The number of values of N contained in the vector NVAL.
77 *> \endverbatim
78 *>
79 *> \param[in] NVAL
80 *> \verbatim
81 *> NVAL is INTEGER array, dimension (NN)
82 *> The values of the matrix column dimension N.
83 *> \endverbatim
84 *>
85 *> \param[in] NNS
86 *> \verbatim
87 *> NNS is INTEGER
88 *> The number of values of NRHS contained in the vector NSVAL.
89 *> \endverbatim
90 *>
91 *> \param[in] NSVAL
92 *> \verbatim
93 *> NSVAL is INTEGER array, dimension (NNS)
94 *> The values of the number of right hand sides NRHS.
95 *> \endverbatim
96 *>
97 *> \param[in] NNB
98 *> \verbatim
99 *> NNB is INTEGER
100 *> The number of values of NB and NX contained in the
101 *> vectors NBVAL and NXVAL. The blocking parameters are used
102 *> in pairs (NB,NX).
103 *> \endverbatim
104 *>
105 *> \param[in] NBVAL
106 *> \verbatim
107 *> NBVAL is INTEGER array, dimension (NNB)
108 *> The values of the blocksize NB.
109 *> \endverbatim
110 *>
111 *> \param[in] NXVAL
112 *> \verbatim
113 *> NXVAL is INTEGER array, dimension (NNB)
114 *> The values of the crossover point NX.
115 *> \endverbatim
116 *>
117 *> \param[in] THRESH
118 *> \verbatim
119 *> THRESH is DOUBLE PRECISION
120 *> The threshold value for the test ratios. A result is
121 *> included in the output file if RESULT >= THRESH. To have
122 *> every test ratio printed, use THRESH = 0.
123 *> \endverbatim
124 *>
125 *> \param[in] TSTERR
126 *> \verbatim
127 *> TSTERR is LOGICAL
128 *> Flag that indicates whether error exits are to be tested.
129 *> \endverbatim
130 *>
131 *> \param[out] A
132 *> \verbatim
133 *> A is DOUBLE PRECISION array, dimension (MMAX*NMAX)
134 *> where MMAX is the maximum value of M in MVAL and NMAX is the
135 *> maximum value of N in NVAL.
136 *> \endverbatim
137 *>
138 *> \param[out] COPYA
139 *> \verbatim
140 *> COPYA is DOUBLE PRECISION array, dimension (MMAX*NMAX)
141 *> \endverbatim
142 *>
143 *> \param[out] B
144 *> \verbatim
145 *> B is DOUBLE PRECISION array, dimension (MMAX*NSMAX)
146 *> where MMAX is the maximum value of M in MVAL and NSMAX is the
147 *> maximum value of NRHS in NSVAL.
148 *> \endverbatim
149 *>
150 *> \param[out] COPYB
151 *> \verbatim
152 *> COPYB is DOUBLE PRECISION array, dimension (MMAX*NSMAX)
153 *> \endverbatim
154 *>
155 *> \param[out] C
156 *> \verbatim
157 *> C is DOUBLE PRECISION array, dimension (MMAX*NSMAX)
158 *> \endverbatim
159 *>
160 *> \param[out] S
161 *> \verbatim
162 *> S is DOUBLE PRECISION array, dimension
163 *> (min(MMAX,NMAX))
164 *> \endverbatim
165 *>
166 *> \param[out] COPYS
167 *> \verbatim
168 *> COPYS is DOUBLE PRECISION array, dimension
169 *> (min(MMAX,NMAX))
170 *> \endverbatim
171 *>
172 *> \param[in] NOUT
173 *> \verbatim
174 *> NOUT is INTEGER
175 *> The unit number for output.
176 *> \endverbatim
177 *
178 * Authors:
179 * ========
180 *
181 *> \author Univ. of Tennessee
182 *> \author Univ. of California Berkeley
183 *> \author Univ. of Colorado Denver
184 *> \author NAG Ltd.
185 *
186 *> \ingroup double_lin
187 *
188 * =====================================================================
189  SUBROUTINE ddrvls( DOTYPE, NM, MVAL, NN, NVAL, NNS, NSVAL, NNB,
190  $ NBVAL, NXVAL, THRESH, TSTERR, A, COPYA, B,
191  $ COPYB, C, S, COPYS, NOUT )
192 *
193 * -- LAPACK test routine --
194 * -- LAPACK is a software package provided by Univ. of Tennessee, --
195 * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
196 *
197 * .. Scalar Arguments ..
198  LOGICAL TSTERR
199  INTEGER NM, NN, NNB, NNS, NOUT
200  DOUBLE PRECISION THRESH
201 * ..
202 * .. Array Arguments ..
203  LOGICAL DOTYPE( * )
204  INTEGER MVAL( * ), NBVAL( * ), NSVAL( * ),
205  $ nval( * ), nxval( * )
206  DOUBLE PRECISION A( * ), B( * ), C( * ), COPYA( * ), COPYB( * ),
207  $ COPYS( * ), S( * )
208 * ..
209 *
210 * =====================================================================
211 *
212 * .. Parameters ..
213  INTEGER NTESTS
214  PARAMETER ( NTESTS = 16 )
215  INTEGER SMLSIZ
216  parameter( smlsiz = 25 )
217  DOUBLE PRECISION ONE, TWO, ZERO
218  parameter( one = 1.0d0, two = 2.0d0, zero = 0.0d0 )
219 * ..
220 * .. Local Scalars ..
221  CHARACTER TRANS
222  CHARACTER*3 PATH
223  INTEGER CRANK, I, IM, IMB, IN, INB, INFO, INS, IRANK,
224  $ iscale, itran, itype, j, k, lda, ldb, ldwork,
225  $ lwlsy, lwork, m, mnmin, n, nb, ncols, nerrs,
226  $ nfail, nrhs, nrows, nrun, rank, mb,
227  $ mmax, nmax, nsmax, liwork,
228  $ lwork_dgels, lwork_dgetsls, lwork_dgelss,
229  $ lwork_dgelsy, lwork_dgelsd
230  DOUBLE PRECISION EPS, NORMA, NORMB, RCOND
231 * ..
232 * .. Local Arrays ..
233  INTEGER ISEED( 4 ), ISEEDY( 4 ), IWQ( 1 )
234  DOUBLE PRECISION RESULT( NTESTS ), WQ( 1 )
235 * ..
236 * .. Allocatable Arrays ..
237  DOUBLE PRECISION, ALLOCATABLE :: WORK (:)
238  INTEGER, ALLOCATABLE :: IWORK (:)
239 * ..
240 * .. External Functions ..
241  DOUBLE PRECISION DASUM, DLAMCH, DQRT12, DQRT14, DQRT17
242  EXTERNAL DASUM, DLAMCH, DQRT12, DQRT14, DQRT17
243 * ..
244 * .. External Subroutines ..
245  EXTERNAL alaerh, alahd, alasvm, daxpy, derrls, dgels,
246  $ dgelsd, dgelss, dgelsy, dgemm, dlacpy,
247  $ dlarnv, dlasrt, dqrt13, dqrt15, dqrt16, dscal,
248  $ xlaenv
249 * ..
250 * .. Intrinsic Functions ..
251  INTRINSIC dble, int, log, max, min, sqrt
252 * ..
253 * .. Scalars in Common ..
254  LOGICAL LERR, OK
255  CHARACTER*32 SRNAMT
256  INTEGER INFOT, IOUNIT
257 * ..
258 * .. Common blocks ..
259  COMMON / infoc / infot, iounit, ok, lerr
260  COMMON / srnamc / srnamt
261 * ..
262 * .. Data statements ..
263  DATA iseedy / 1988, 1989, 1990, 1991 /
264 * ..
265 * .. Executable Statements ..
266 *
267 * Initialize constants and the random number seed.
268 *
269  path( 1: 1 ) = 'Double precision'
270  path( 2: 3 ) = 'LS'
271  nrun = 0
272  nfail = 0
273  nerrs = 0
274  DO 10 i = 1, 4
275  iseed( i ) = iseedy( i )
276  10 CONTINUE
277  eps = dlamch( 'Epsilon' )
278 *
279 * Threshold for rank estimation
280 *
281  rcond = sqrt( eps ) - ( sqrt( eps )-eps ) / 2
282 *
283 * Test the error exits
284 *
285  CALL xlaenv( 2, 2 )
286  CALL xlaenv( 9, smlsiz )
287  IF( tsterr )
288  $ CALL derrls( path, nout )
289 *
290 * Print the header if NM = 0 or NN = 0 and THRESH = 0.
291 *
292  IF( ( nm.EQ.0 .OR. nn.EQ.0 ) .AND. thresh.EQ.zero )
293  $ CALL alahd( nout, path )
294  infot = 0
295  CALL xlaenv( 2, 2 )
296  CALL xlaenv( 9, smlsiz )
297 *
298 * Compute maximal workspace needed for all routines
299 *
300  nmax = 0
301  mmax = 0
302  nsmax = 0
303  DO i = 1, nm
304  IF ( mval( i ).GT.mmax ) THEN
305  mmax = mval( i )
306  END IF
307  ENDDO
308  DO i = 1, nn
309  IF ( nval( i ).GT.nmax ) THEN
310  nmax = nval( i )
311  END IF
312  ENDDO
313  DO i = 1, nns
314  IF ( nsval( i ).GT.nsmax ) THEN
315  nsmax = nsval( i )
316  END IF
317  ENDDO
318  m = mmax
319  n = nmax
320  nrhs = nsmax
321  mnmin = max( min( m, n ), 1 )
322 *
323 * Compute workspace needed for routines
324 * DQRT14, DQRT17 (two side cases), DQRT15 and DQRT12
325 *
326  lwork = max( 1, ( m+n )*nrhs,
327  $ ( n+nrhs )*( m+2 ), ( m+nrhs )*( n+2 ),
328  $ max( m+mnmin, nrhs*mnmin,2*n+m ),
329  $ max( m*n+4*mnmin+max(m,n), m*n+2*mnmin+4*n ) )
330  liwork = 1
331 *
332 * Iterate through all test cases and compute necessary workspace
333 * sizes for ?GELS, ?GETSLS, ?GELSY, ?GELSS and ?GELSD routines.
334 *
335  DO im = 1, nm
336  m = mval( im )
337  lda = max( 1, m )
338  DO in = 1, nn
339  n = nval( in )
340  mnmin = max(min( m, n ),1)
341  ldb = max( 1, m, n )
342  DO ins = 1, nns
343  nrhs = nsval( ins )
344  DO irank = 1, 2
345  DO iscale = 1, 3
346  itype = ( irank-1 )*3 + iscale
347  IF( dotype( itype ) ) THEN
348  IF( irank.EQ.1 ) THEN
349  DO itran = 1, 2
350  IF( itran.EQ.1 ) THEN
351  trans = 'N'
352  ELSE
353  trans = 'T'
354  END IF
355 *
356 * Compute workspace needed for DGELS
357  CALL dgels( trans, m, n, nrhs, a, lda,
358  $ b, ldb, wq, -1, info )
359  lwork_dgels = int( wq( 1 ) )
360 * Compute workspace needed for DGETSLS
361  CALL dgetsls( trans, m, n, nrhs, a, lda,
362  $ b, ldb, wq, -1, info )
363  lwork_dgetsls = int( wq( 1 ) )
364  ENDDO
365  END IF
366 * Compute workspace needed for DGELSY
367  CALL dgelsy( m, n, nrhs, a, lda, b, ldb, iwq,
368  $ rcond, crank, wq, -1, info )
369  lwork_dgelsy = int( wq( 1 ) )
370 * Compute workspace needed for DGELSS
371  CALL dgelss( m, n, nrhs, a, lda, b, ldb, s,
372  $ rcond, crank, wq, -1 , info )
373  lwork_dgelss = int( wq( 1 ) )
374 * Compute workspace needed for DGELSD
375  CALL dgelsd( m, n, nrhs, a, lda, b, ldb, s,
376  $ rcond, crank, wq, -1, iwq, info )
377  lwork_dgelsd = int( wq( 1 ) )
378 * Compute LIWORK workspace needed for DGELSY and DGELSD
379  liwork = max( liwork, n, iwq( 1 ) )
380 * Compute LWORK workspace needed for all functions
381  lwork = max( lwork, lwork_dgels, lwork_dgetsls,
382  $ lwork_dgelsy, lwork_dgelss,
383  $ lwork_dgelsd )
384  END IF
385  ENDDO
386  ENDDO
387  ENDDO
388  ENDDO
389  ENDDO
390 *
391  lwlsy = lwork
392 *
393  ALLOCATE( work( lwork ) )
394  ALLOCATE( iwork( liwork ) )
395 *
396  DO 150 im = 1, nm
397  m = mval( im )
398  lda = max( 1, m )
399 *
400  DO 140 in = 1, nn
401  n = nval( in )
402  mnmin = max(min( m, n ),1)
403  ldb = max( 1, m, n )
404  mb = (mnmin+1)
405 *
406  DO 130 ins = 1, nns
407  nrhs = nsval( ins )
408 *
409  DO 120 irank = 1, 2
410  DO 110 iscale = 1, 3
411  itype = ( irank-1 )*3 + iscale
412  IF( .NOT.dotype( itype ) )
413  $ GO TO 110
414 *
415  IF( irank.EQ.1 ) THEN
416 *
417 * Test DGELS
418 *
419 * Generate a matrix of scaling type ISCALE
420 *
421  CALL dqrt13( iscale, m, n, copya, lda, norma,
422  $ iseed )
423  DO 40 inb = 1, nnb
424  nb = nbval( inb )
425  CALL xlaenv( 1, nb )
426  CALL xlaenv( 3, nxval( inb ) )
427 *
428  DO 30 itran = 1, 2
429  IF( itran.EQ.1 ) THEN
430  trans = 'N'
431  nrows = m
432  ncols = n
433  ELSE
434  trans = 'T'
435  nrows = n
436  ncols = m
437  END IF
438  ldwork = max( 1, ncols )
439 *
440 * Set up a consistent rhs
441 *
442  IF( ncols.GT.0 ) THEN
443  CALL dlarnv( 2, iseed, ncols*nrhs,
444  $ work )
445  CALL dscal( ncols*nrhs,
446  $ one / dble( ncols ), work,
447  $ 1 )
448  END IF
449  CALL dgemm( trans, 'No transpose', nrows,
450  $ nrhs, ncols, one, copya, lda,
451  $ work, ldwork, zero, b, ldb )
452  CALL dlacpy( 'Full', nrows, nrhs, b, ldb,
453  $ copyb, ldb )
454 *
455 * Solve LS or overdetermined system
456 *
457  IF( m.GT.0 .AND. n.GT.0 ) THEN
458  CALL dlacpy( 'Full', m, n, copya, lda,
459  $ a, lda )
460  CALL dlacpy( 'Full', nrows, nrhs,
461  $ copyb, ldb, b, ldb )
462  END IF
463  srnamt = 'DGELS '
464  CALL dgels( trans, m, n, nrhs, a, lda, b,
465  $ ldb, work, lwork, info )
466  IF( info.NE.0 )
467  $ CALL alaerh( path, 'DGELS ', info, 0,
468  $ trans, m, n, nrhs, -1, nb,
469  $ itype, nfail, nerrs,
470  $ nout )
471 *
472 * Check correctness of results
473 *
474  ldwork = max( 1, nrows )
475  IF( nrows.GT.0 .AND. nrhs.GT.0 )
476  $ CALL dlacpy( 'Full', nrows, nrhs,
477  $ copyb, ldb, c, ldb )
478  CALL dqrt16( trans, m, n, nrhs, copya,
479  $ lda, b, ldb, c, ldb, work,
480  $ result( 1 ) )
481 *
482  IF( ( itran.EQ.1 .AND. m.GE.n ) .OR.
483  $ ( itran.EQ.2 .AND. m.LT.n ) ) THEN
484 *
485 * Solving LS system
486 *
487  result( 2 ) = dqrt17( trans, 1, m, n,
488  $ nrhs, copya, lda, b, ldb,
489  $ copyb, ldb, c, work,
490  $ lwork )
491  ELSE
492 *
493 * Solving overdetermined system
494 *
495  result( 2 ) = dqrt14( trans, m, n,
496  $ nrhs, copya, lda, b, ldb,
497  $ work, lwork )
498  END IF
499 *
500 * Print information about the tests that
501 * did not pass the threshold.
502 *
503  DO 20 k = 1, 2
504  IF( result( k ).GE.thresh ) THEN
505  IF( nfail.EQ.0 .AND. nerrs.EQ.0 )
506  $ CALL alahd( nout, path )
507  WRITE( nout, fmt = 9999 )trans, m,
508  $ n, nrhs, nb, itype, k,
509  $ result( k )
510  nfail = nfail + 1
511  END IF
512  20 CONTINUE
513  nrun = nrun + 2
514  30 CONTINUE
515  40 CONTINUE
516 *
517 *
518 * Test DGETSLS
519 *
520 * Generate a matrix of scaling type ISCALE
521 *
522  CALL dqrt13( iscale, m, n, copya, lda, norma,
523  $ iseed )
524  DO 65 inb = 1, nnb
525  mb = nbval( inb )
526  CALL xlaenv( 1, mb )
527  DO 62 imb = 1, nnb
528  nb = nbval( imb )
529  CALL xlaenv( 2, nb )
530 *
531  DO 60 itran = 1, 2
532  IF( itran.EQ.1 ) THEN
533  trans = 'N'
534  nrows = m
535  ncols = n
536  ELSE
537  trans = 'T'
538  nrows = n
539  ncols = m
540  END IF
541  ldwork = max( 1, ncols )
542 *
543 * Set up a consistent rhs
544 *
545  IF( ncols.GT.0 ) THEN
546  CALL dlarnv( 2, iseed, ncols*nrhs,
547  $ work )
548  CALL dscal( ncols*nrhs,
549  $ one / dble( ncols ), work,
550  $ 1 )
551  END IF
552  CALL dgemm( trans, 'No transpose', nrows,
553  $ nrhs, ncols, one, copya, lda,
554  $ work, ldwork, zero, b, ldb )
555  CALL dlacpy( 'Full', nrows, nrhs, b, ldb,
556  $ copyb, ldb )
557 *
558 * Solve LS or overdetermined system
559 *
560  IF( m.GT.0 .AND. n.GT.0 ) THEN
561  CALL dlacpy( 'Full', m, n, copya, lda,
562  $ a, lda )
563  CALL dlacpy( 'Full', nrows, nrhs,
564  $ copyb, ldb, b, ldb )
565  END IF
566  srnamt = 'DGETSLS '
567  CALL dgetsls( trans, m, n, nrhs, a,
568  $ lda, b, ldb, work, lwork, info )
569  IF( info.NE.0 )
570  $ CALL alaerh( path, 'DGETSLS ', info, 0,
571  $ trans, m, n, nrhs, -1, nb,
572  $ itype, nfail, nerrs,
573  $ nout )
574 *
575 * Check correctness of results
576 *
577  ldwork = max( 1, nrows )
578  IF( nrows.GT.0 .AND. nrhs.GT.0 )
579  $ CALL dlacpy( 'Full', nrows, nrhs,
580  $ copyb, ldb, c, ldb )
581  CALL dqrt16( trans, m, n, nrhs, copya,
582  $ lda, b, ldb, c, ldb, work,
583  $ result( 15 ) )
584 *
585  IF( ( itran.EQ.1 .AND. m.GE.n ) .OR.
586  $ ( itran.EQ.2 .AND. m.LT.n ) ) THEN
587 *
588 * Solving LS system
589 *
590  result( 16 ) = dqrt17( trans, 1, m, n,
591  $ nrhs, copya, lda, b, ldb,
592  $ copyb, ldb, c, work,
593  $ lwork )
594  ELSE
595 *
596 * Solving overdetermined system
597 *
598  result( 16 ) = dqrt14( trans, m, n,
599  $ nrhs, copya, lda, b, ldb,
600  $ work, lwork )
601  END IF
602 *
603 * Print information about the tests that
604 * did not pass the threshold.
605 *
606  DO 50 k = 15, 16
607  IF( result( k ).GE.thresh ) THEN
608  IF( nfail.EQ.0 .AND. nerrs.EQ.0 )
609  $ CALL alahd( nout, path )
610  WRITE( nout, fmt = 9997 )trans, m,
611  $ n, nrhs, mb, nb, itype, k,
612  $ result( k )
613  nfail = nfail + 1
614  END IF
615  50 CONTINUE
616  nrun = nrun + 2
617  60 CONTINUE
618  62 CONTINUE
619  65 CONTINUE
620  END IF
621 *
622 * Generate a matrix of scaling type ISCALE and rank
623 * type IRANK.
624 *
625  CALL dqrt15( iscale, irank, m, n, nrhs, copya, lda,
626  $ copyb, ldb, copys, rank, norma, normb,
627  $ iseed, work, lwork )
628 *
629 * workspace used: MAX(M+MIN(M,N),NRHS*MIN(M,N),2*N+M)
630 *
631  ldwork = max( 1, m )
632 *
633 * Loop for testing different block sizes.
634 *
635  DO 100 inb = 1, nnb
636  nb = nbval( inb )
637  CALL xlaenv( 1, nb )
638  CALL xlaenv( 3, nxval( inb ) )
639 *
640 * Test DGELSY
641 *
642 * DGELSY: Compute the minimum-norm solution X
643 * to min( norm( A * X - B ) )
644 * using the rank-revealing orthogonal
645 * factorization.
646 *
647 * Initialize vector IWORK.
648 *
649  DO 70 j = 1, n
650  iwork( j ) = 0
651  70 CONTINUE
652 *
653  CALL dlacpy( 'Full', m, n, copya, lda, a, lda )
654  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, b,
655  $ ldb )
656 *
657  srnamt = 'DGELSY'
658  CALL dgelsy( m, n, nrhs, a, lda, b, ldb, iwork,
659  $ rcond, crank, work, lwlsy, info )
660  IF( info.NE.0 )
661  $ CALL alaerh( path, 'DGELSY', info, 0, ' ', m,
662  $ n, nrhs, -1, nb, itype, nfail,
663  $ nerrs, nout )
664 *
665 * Test 3: Compute relative error in svd
666 * workspace: M*N + 4*MIN(M,N) + MAX(M,N)
667 *
668  result( 3 ) = dqrt12( crank, crank, a, lda,
669  $ copys, work, lwork )
670 *
671 * Test 4: Compute error in solution
672 * workspace: M*NRHS + M
673 *
674  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, work,
675  $ ldwork )
676  CALL dqrt16( 'No transpose', m, n, nrhs, copya,
677  $ lda, b, ldb, work, ldwork,
678  $ work( m*nrhs+1 ), result( 4 ) )
679 *
680 * Test 5: Check norm of r'*A
681 * workspace: NRHS*(M+N)
682 *
683  result( 5 ) = zero
684  IF( m.GT.crank )
685  $ result( 5 ) = dqrt17( 'No transpose', 1, m,
686  $ n, nrhs, copya, lda, b, ldb,
687  $ copyb, ldb, c, work, lwork )
688 *
689 * Test 6: Check if x is in the rowspace of A
690 * workspace: (M+NRHS)*(N+2)
691 *
692  result( 6 ) = zero
693 *
694  IF( n.GT.crank )
695  $ result( 6 ) = dqrt14( 'No transpose', m, n,
696  $ nrhs, copya, lda, b, ldb,
697  $ work, lwork )
698 *
699 * Test DGELSS
700 *
701 * DGELSS: Compute the minimum-norm solution X
702 * to min( norm( A * X - B ) )
703 * using the SVD.
704 *
705  CALL dlacpy( 'Full', m, n, copya, lda, a, lda )
706  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, b,
707  $ ldb )
708  srnamt = 'DGELSS'
709  CALL dgelss( m, n, nrhs, a, lda, b, ldb, s,
710  $ rcond, crank, work, lwork, info )
711  IF( info.NE.0 )
712  $ CALL alaerh( path, 'DGELSS', info, 0, ' ', m,
713  $ n, nrhs, -1, nb, itype, nfail,
714  $ nerrs, nout )
715 *
716 * workspace used: 3*min(m,n) +
717 * max(2*min(m,n),nrhs,max(m,n))
718 *
719 * Test 7: Compute relative error in svd
720 *
721  IF( rank.GT.0 ) THEN
722  CALL daxpy( mnmin, -one, copys, 1, s, 1 )
723  result( 7 ) = dasum( mnmin, s, 1 ) /
724  $ dasum( mnmin, copys, 1 ) /
725  $ ( eps*dble( mnmin ) )
726  ELSE
727  result( 7 ) = zero
728  END IF
729 *
730 * Test 8: Compute error in solution
731 *
732  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, work,
733  $ ldwork )
734  CALL dqrt16( 'No transpose', m, n, nrhs, copya,
735  $ lda, b, ldb, work, ldwork,
736  $ work( m*nrhs+1 ), result( 8 ) )
737 *
738 * Test 9: Check norm of r'*A
739 *
740  result( 9 ) = zero
741  IF( m.GT.crank )
742  $ result( 9 ) = dqrt17( 'No transpose', 1, m,
743  $ n, nrhs, copya, lda, b, ldb,
744  $ copyb, ldb, c, work, lwork )
745 *
746 * Test 10: Check if x is in the rowspace of A
747 *
748  result( 10 ) = zero
749  IF( n.GT.crank )
750  $ result( 10 ) = dqrt14( 'No transpose', m, n,
751  $ nrhs, copya, lda, b, ldb,
752  $ work, lwork )
753 *
754 * Test DGELSD
755 *
756 * DGELSD: Compute the minimum-norm solution X
757 * to min( norm( A * X - B ) ) using a
758 * divide and conquer SVD.
759 *
760 * Initialize vector IWORK.
761 *
762  DO 80 j = 1, n
763  iwork( j ) = 0
764  80 CONTINUE
765 *
766  CALL dlacpy( 'Full', m, n, copya, lda, a, lda )
767  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, b,
768  $ ldb )
769 *
770  srnamt = 'DGELSD'
771  CALL dgelsd( m, n, nrhs, a, lda, b, ldb, s,
772  $ rcond, crank, work, lwork, iwork,
773  $ info )
774  IF( info.NE.0 )
775  $ CALL alaerh( path, 'DGELSD', info, 0, ' ', m,
776  $ n, nrhs, -1, nb, itype, nfail,
777  $ nerrs, nout )
778 *
779 * Test 11: Compute relative error in svd
780 *
781  IF( rank.GT.0 ) THEN
782  CALL daxpy( mnmin, -one, copys, 1, s, 1 )
783  result( 11 ) = dasum( mnmin, s, 1 ) /
784  $ dasum( mnmin, copys, 1 ) /
785  $ ( eps*dble( mnmin ) )
786  ELSE
787  result( 11 ) = zero
788  END IF
789 *
790 * Test 12: Compute error in solution
791 *
792  CALL dlacpy( 'Full', m, nrhs, copyb, ldb, work,
793  $ ldwork )
794  CALL dqrt16( 'No transpose', m, n, nrhs, copya,
795  $ lda, b, ldb, work, ldwork,
796  $ work( m*nrhs+1 ), result( 12 ) )
797 *
798 * Test 13: Check norm of r'*A
799 *
800  result( 13 ) = zero
801  IF( m.GT.crank )
802  $ result( 13 ) = dqrt17( 'No transpose', 1, m,
803  $ n, nrhs, copya, lda, b, ldb,
804  $ copyb, ldb, c, work, lwork )
805 *
806 * Test 14: Check if x is in the rowspace of A
807 *
808  result( 14 ) = zero
809  IF( n.GT.crank )
810  $ result( 14 ) = dqrt14( 'No transpose', m, n,
811  $ nrhs, copya, lda, b, ldb,
812  $ work, lwork )
813 *
814 * Print information about the tests that did not
815 * pass the threshold.
816 *
817  DO 90 k = 3, 14
818  IF( result( k ).GE.thresh ) THEN
819  IF( nfail.EQ.0 .AND. nerrs.EQ.0 )
820  $ CALL alahd( nout, path )
821  WRITE( nout, fmt = 9998 )m, n, nrhs, nb,
822  $ itype, k, result( k )
823  nfail = nfail + 1
824  END IF
825  90 CONTINUE
826  nrun = nrun + 12
827 *
828  100 CONTINUE
829  110 CONTINUE
830  120 CONTINUE
831  130 CONTINUE
832  140 CONTINUE
833  150 CONTINUE
834 *
835 * Print a summary of the results.
836 *
837  CALL alasvm( path, nout, nfail, nrun, nerrs )
838 *
839  9999 FORMAT( ' TRANS=''', a1, ''', M=', i5, ', N=', i5, ', NRHS=', i4,
840  $ ', NB=', i4, ', type', i2, ', test(', i2, ')=', g12.5 )
841  9998 FORMAT( ' M=', i5, ', N=', i5, ', NRHS=', i4, ', NB=', i4,
842  $ ', type', i2, ', test(', i2, ')=', g12.5 )
843  9997 FORMAT( ' TRANS=''', a1,' M=', i5, ', N=', i5, ', NRHS=', i4,
844  $ ', MB=', i4,', NB=', i4,', type', i2,
845  $ ', test(', i2, ')=', g12.5 )
846 *
847  DEALLOCATE( work )
848  DEALLOCATE( iwork )
849  RETURN
850 *
851 * End of DDRVLS
852 *
853  END
dlarnv
subroutine dlarnv(IDIST, ISEED, N, X)
DLARNV returns a vector of random numbers from a uniform or normal distribution.
Definition: dlarnv.f:97
dlacpy
subroutine dlacpy(UPLO, M, N, A, LDA, B, LDB)
DLACPY copies all or part of one two-dimensional array to another.
Definition: dlacpy.f:103
alasvm
subroutine alasvm(TYPE, NOUT, NFAIL, NRUN, NERRS)
ALASVM
Definition: alasvm.f:73
xlaenv
subroutine xlaenv(ISPEC, NVALUE)
XLAENV
Definition: xlaenv.f:81
alahd
subroutine alahd(IOUNIT, PATH)
ALAHD
Definition: alahd.f:107
alaerh
subroutine alaerh(PATH, SUBNAM, INFO, INFOE, OPTS, M, N, KL, KU, N5, IMAT, NFAIL, NERRS, NOUT)
ALAERH
Definition: alaerh.f:147
dlasrt
subroutine dlasrt(ID, N, D, INFO)
DLASRT sorts numbers in increasing or decreasing order.
Definition: dlasrt.f:88
dscal
subroutine dscal(N, DA, DX, INCX)
DSCAL
Definition: dscal.f:79
daxpy
subroutine daxpy(N, DA, DX, INCX, DY, INCY)
DAXPY
Definition: daxpy.f:89
dgemm
subroutine dgemm(TRANSA, TRANSB, M, N, K, ALPHA, A, LDA, B, LDB, BETA, C, LDC)
DGEMM
Definition: dgemm.f:187
derrls
subroutine derrls(PATH, NUNIT)
DERRLS
Definition: derrls.f:55
dqrt15
subroutine dqrt15(SCALE, RKSEL, M, N, NRHS, A, LDA, B, LDB, S, RANK, NORMA, NORMB, ISEED, WORK, LWORK)
DQRT15
Definition: dqrt15.f:148
ddrvls
subroutine ddrvls(DOTYPE, NM, MVAL, NN, NVAL, NNS, NSVAL, NNB, NBVAL, NXVAL, THRESH, TSTERR, A, COPYA, B, COPYB, C, S, COPYS, NOUT)
DDRVLS
Definition: ddrvls.f:192
dqrt13
subroutine dqrt13(SCALE, M, N, A, LDA, NORMA, ISEED)
DQRT13
Definition: dqrt13.f:91
dqrt16
subroutine dqrt16(TRANS, M, N, NRHS, A, LDA, X, LDX, B, LDB, RWORK, RESID)
DQRT16
Definition: dqrt16.f:133
dgels
subroutine dgels(TRANS, M, N, NRHS, A, LDA, B, LDB, WORK, LWORK, INFO)
DGELS solves overdetermined or underdetermined systems for GE matrices
Definition: dgels.f:183
dgetsls
subroutine dgetsls(TRANS, M, N, NRHS, A, LDA, B, LDB, WORK, LWORK, INFO)
DGETSLS
Definition: dgetsls.f:162
dgelsy
subroutine dgelsy(M, N, NRHS, A, LDA, B, LDB, JPVT, RCOND, RANK, WORK, LWORK, INFO)
DGELSY solves overdetermined or underdetermined systems for GE matrices
Definition: dgelsy.f:204
dgelsd
subroutine dgelsd(M, N, NRHS, A, LDA, B, LDB, S, RCOND, RANK, WORK, LWORK, IWORK, INFO)
DGELSD computes the minimum-norm solution to a linear least squares problem for GE matrices
Definition: dgelsd.f:209
dgelss
subroutine dgelss(M, N, NRHS, A, LDA, B, LDB, S, RCOND, RANK, WORK, LWORK, INFO)
DGELSS solves overdetermined or underdetermined systems for GE matrices
Definition: dgelss.f:172