phynexis-deps/lapack
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Files
introduce-cpp
lapack/TESTING/LIN/ddrvrf4.f
T
Julie e18d437924 Making LAPACK's code eternal... no more version and date in source files. 
GitHub is now enabling us to track accurately version and date.
No need for this anymore.
2021-03-25 10:16:58 -07:00

358 lines
11 KiB
FortranFixed
Raw Permalink Blame History

*> \brief \b DDRVRF4
*
* =========== DOCUMENTATION ===========
*
* Online html documentation available at
* http://www.netlib.org/lapack/explore-html/
*
* Definition:
* ===========
*
* SUBROUTINE DDRVRF4( NOUT, NN, NVAL, THRESH, C1, C2, LDC, CRF, A,
* + LDA, D_WORK_DLANGE )
*
* .. Scalar Arguments ..
* INTEGER LDA, LDC, NN, NOUT
* DOUBLE PRECISION THRESH
* ..
* .. Array Arguments ..
* INTEGER NVAL( NN )
* DOUBLE PRECISION A( LDA, * ), C1( LDC, * ), C2( LDC, *),
* + CRF( * ), D_WORK_DLANGE( * )
* ..
*
*
*> \par Purpose:
* =============
*>
*> \verbatim
*>
*> DDRVRF4 tests the LAPACK RFP routines:
*> DSFRK
*> \endverbatim
*
* Arguments:
* ==========
*
*> \param[in] NOUT
*> \verbatim
*> NOUT is INTEGER
*> The unit number for output.
*> \endverbatim
*>
*> \param[in] NN
*> \verbatim
*> NN is INTEGER
*> The number of values of N contained in the vector NVAL.
*> \endverbatim
*>
*> \param[in] NVAL
*> \verbatim
*> NVAL is INTEGER array, dimension (NN)
*> The values of the matrix dimension N.
*> \endverbatim
*>
*> \param[in] THRESH
*> \verbatim
*> THRESH is DOUBLE PRECISION
*> The threshold value for the test ratios. A result is
*> included in the output file if RESULT >= THRESH. To
*> have every test ratio printed, use THRESH = 0.
*> \endverbatim
*>
*> \param[out] C1
*> \verbatim
*> C1 is DOUBLE PRECISION array,
*> dimension (LDC,NMAX)
*> \endverbatim
*>
*> \param[out] C2
*> \verbatim
*> C2 is DOUBLE PRECISION array,
*> dimension (LDC,NMAX)
*> \endverbatim
*>
*> \param[in] LDC
*> \verbatim
*> LDC is INTEGER
*> The leading dimension of the array A.
*> LDA >= max(1,NMAX).
*> \endverbatim
*>
*> \param[out] CRF
*> \verbatim
*> CRF is DOUBLE PRECISION array,
*> dimension ((NMAX*(NMAX+1))/2).
*> \endverbatim
*>
*> \param[out] A
*> \verbatim
*> A is DOUBLE PRECISION array,
*> dimension (LDA,NMAX)
*> \endverbatim
*>
*> \param[in] LDA
*> \verbatim
*> LDA is INTEGER
*> The leading dimension of the array A. LDA >= max(1,NMAX).
*> \endverbatim
*>
*> \param[out] D_WORK_DLANGE
*> \verbatim
*> D_WORK_DLANGE is DOUBLE PRECISION array, dimension (NMAX)
*> \endverbatim
*
* Authors:
* ========
*
*> \author Univ. of Tennessee
*> \author Univ. of California Berkeley
*> \author Univ. of Colorado Denver
*> \author NAG Ltd.
*
*> \ingroup double_lin
*
* =====================================================================
SUBROUTINE DDRVRF4( NOUT, NN, NVAL, THRESH, C1, C2, LDC, CRF, A,
+ LDA, D_WORK_DLANGE )
*
* -- LAPACK test routine --
* -- LAPACK is a software package provided by Univ. of Tennessee, --
* -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
*
* .. Scalar Arguments ..
INTEGER LDA, LDC, NN, NOUT
DOUBLE PRECISION THRESH
* ..
* .. Array Arguments ..
INTEGER NVAL( NN )
DOUBLE PRECISION A( LDA, * ), C1( LDC, * ), C2( LDC, *),
+ CRF( * ), D_WORK_DLANGE( * )
* ..
*
* =====================================================================
* ..
* .. Parameters ..
DOUBLE PRECISION ZERO, ONE
PARAMETER ( ZERO = 0.0D+0, ONE = 1.0D+0 )
INTEGER NTESTS
PARAMETER ( NTESTS = 1 )
* ..
* .. Local Scalars ..
CHARACTER UPLO, CFORM, TRANS
INTEGER I, IFORM, IIK, IIN, INFO, IUPLO, J, K, N,
+ NFAIL, NRUN, IALPHA, ITRANS
DOUBLE PRECISION ALPHA, BETA, EPS, NORMA, NORMC
* ..
* .. Local Arrays ..
CHARACTER UPLOS( 2 ), FORMS( 2 ), TRANSS( 2 )
INTEGER ISEED( 4 ), ISEEDY( 4 )
DOUBLE PRECISION RESULT( NTESTS )
* ..
* .. External Functions ..
DOUBLE PRECISION DLAMCH, DLARND, DLANGE
EXTERNAL DLAMCH, DLARND, DLANGE
* ..
* .. External Subroutines ..
EXTERNAL DSYRK, DSFRK, DTFTTR, DTRTTF
* ..
* .. Intrinsic Functions ..
INTRINSIC ABS, MAX
* ..
* .. Scalars in Common ..
CHARACTER*32 SRNAMT
* ..
* .. Common blocks ..
COMMON / SRNAMC / SRNAMT
* ..
* .. Data statements ..
DATA ISEEDY / 1988, 1989, 1990, 1991 /
DATA UPLOS / 'U', 'L' /
DATA FORMS / 'N', 'T' /
DATA TRANSS / 'N', 'T' /
* ..
* .. Executable Statements ..
*
* Initialize constants and the random number seed.
*
NRUN = 0
NFAIL = 0
INFO = 0
DO 10 I = 1, 4
ISEED( I ) = ISEEDY( I )
10 CONTINUE
EPS = DLAMCH( 'Precision' )
*
DO 150 IIN = 1, NN
*
N = NVAL( IIN )
*
DO 140 IIK = 1, NN
*
K = NVAL( IIN )
*
DO 130 IFORM = 1, 2
*
CFORM = FORMS( IFORM )
*
DO 120 IUPLO = 1, 2
*
UPLO = UPLOS( IUPLO )
*
DO 110 ITRANS = 1, 2
*
TRANS = TRANSS( ITRANS )
*
DO 100 IALPHA = 1, 4
*
IF ( IALPHA.EQ. 1) THEN
ALPHA = ZERO
BETA = ZERO
ELSE IF ( IALPHA.EQ. 2) THEN
ALPHA = ONE
BETA = ZERO
ELSE IF ( IALPHA.EQ. 3) THEN
ALPHA = ZERO
BETA = ONE
ELSE
ALPHA = DLARND( 2, ISEED )
BETA = DLARND( 2, ISEED )
END IF
*
* All the parameters are set:
* CFORM, UPLO, TRANS, M, N,
* ALPHA, and BETA
* READY TO TEST!
*
NRUN = NRUN + 1
*
IF ( ITRANS.EQ.1 ) THEN
*
* In this case we are NOTRANS, so A is N-by-K
*
DO J = 1, K
DO I = 1, N
A( I, J) = DLARND( 2, ISEED )
END DO
END DO
*
NORMA = DLANGE( 'I', N, K, A, LDA,
+ D_WORK_DLANGE )
*
ELSE
*
* In this case we are TRANS, so A is K-by-N
*
DO J = 1,N
DO I = 1, K
A( I, J) = DLARND( 2, ISEED )
END DO
END DO
*
NORMA = DLANGE( 'I', K, N, A, LDA,
+ D_WORK_DLANGE )
*
END IF
*
* Generate C1 our N--by--N symmetric matrix.
* Make sure C2 has the same upper/lower part,
* (the one that we do not touch), so
* copy the initial C1 in C2 in it.
*
DO J = 1, N
DO I = 1, N
C1( I, J) = DLARND( 2, ISEED )
C2(I,J) = C1(I,J)
END DO
END DO
*
* (See comment later on for why we use DLANGE and
* not DLANSY for C1.)
*
NORMC = DLANGE( 'I', N, N, C1, LDC,
+ D_WORK_DLANGE )
*
SRNAMT = 'DTRTTF'
CALL DTRTTF( CFORM, UPLO, N, C1, LDC, CRF,
+ INFO )
*
* call dsyrk the BLAS routine -> gives C1
*
SRNAMT = 'DSYRK '
CALL DSYRK( UPLO, TRANS, N, K, ALPHA, A, LDA,
+ BETA, C1, LDC )
*
* call dsfrk the RFP routine -> gives CRF
*
SRNAMT = 'DSFRK '
CALL DSFRK( CFORM, UPLO, TRANS, N, K, ALPHA, A,
+ LDA, BETA, CRF )
*
* convert CRF in full format -> gives C2
*
SRNAMT = 'DTFTTR'
CALL DTFTTR( CFORM, UPLO, N, CRF, C2, LDC,
+ INFO )
*
* compare C1 and C2
*
DO J = 1, N
DO I = 1, N
C1(I,J) = C1(I,J)-C2(I,J)
END DO
END DO
*
* Yes, C1 is symmetric so we could call DLANSY,
* but we want to check the upper part that is
* supposed to be unchanged and the diagonal that
* is supposed to be real -> DLANGE
*
RESULT(1) = DLANGE( 'I', N, N, C1, LDC,
+ D_WORK_DLANGE )
RESULT(1) = RESULT(1)
+ / MAX( ABS( ALPHA ) * NORMA
+ + ABS( BETA ) , ONE )
+ / MAX( N , 1 ) / EPS
*
IF( RESULT(1).GE.THRESH ) THEN
IF( NFAIL.EQ.0 ) THEN
WRITE( NOUT, * )
WRITE( NOUT, FMT = 9999 )
END IF
WRITE( NOUT, FMT = 9997 ) 'DSFRK',
+ CFORM, UPLO, TRANS, N, K, RESULT(1)
NFAIL = NFAIL + 1
END IF
*
100 CONTINUE
110 CONTINUE
120 CONTINUE
130 CONTINUE
140 CONTINUE
150 CONTINUE
*
* Print a summary of the results.
*
IF ( NFAIL.EQ.0 ) THEN
WRITE( NOUT, FMT = 9996 ) 'DSFRK', NRUN
ELSE
WRITE( NOUT, FMT = 9995 ) 'DSFRK', NFAIL, NRUN
END IF
*
9999 FORMAT( 1X, ' *** Error(s) or Failure(s) while testing DSFRK
+ ***')
9997 FORMAT( 1X, ' Failure in ',A5,', CFORM=''',A1,''',',
+ ' UPLO=''',A1,''',',' TRANS=''',A1,''',', ' N=',I3,', K =', I3,
+ ', test=',G12.5)
9996 FORMAT( 1X, 'All tests for ',A5,' auxiliary routine passed the ',
+ 'threshold ( ',I5,' tests run)')
9995 FORMAT( 1X, A6, ' auxiliary routine: ',I5,' out of ',I5,
+ ' tests failed to pass the threshold')
*
RETURN
*
* End of DDRVRF4
*
END
Reference in New Issue View Git Blame Copy Permalink