LAPACK: SRC/dpotrf.f Source File

Go to the documentation of this file.
00001       SUBROUTINE DPOTRF( UPLO, N, A, LDA, INFO )
00002 *
00003 *  -- LAPACK routine (version 3.2) --
00004 *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
00005 *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
00006 *     November 2006
00007 *
00008 *     .. Scalar Arguments ..
00009       CHARACTER          UPLO
00010       INTEGER            INFO, LDA, N
00011 *     ..
00012 *     .. Array Arguments ..
00013       DOUBLE PRECISION   A( LDA, * )
00014 *     ..
00015 *
00016 *  Purpose
00017 *  =======
00018 *
00019 *  DPOTRF computes the Cholesky factorization of a real symmetric
00020 *  positive definite matrix A.
00021 *
00022 *  The factorization has the form
00023 *     A = U**T * U,  if UPLO = 'U', or
00024 *     A = L  * L**T,  if UPLO = 'L',
00025 *  where U is an upper triangular matrix and L is lower triangular.
00026 *
00027 *  This is the block version of the algorithm, calling Level 3 BLAS.
00028 *
00029 *  Arguments
00030 *  =========
00031 *
00032 *  UPLO    (input) CHARACTER*1
00033 *          = 'U':  Upper triangle of A is stored;
00034 *          = 'L':  Lower triangle of A is stored.
00035 *
00036 *  N       (input) INTEGER
00037 *          The order of the matrix A.  N >= 0.
00038 *
00039 *  A       (input/output) DOUBLE PRECISION array, dimension (LDA,N)
00040 *          On entry, the symmetric matrix A.  If UPLO = 'U', the leading
00041 *          N-by-N upper triangular part of A contains the upper
00042 *          triangular part of the matrix A, and the strictly lower
00043 *          triangular part of A is not referenced.  If UPLO = 'L', the
00044 *          leading N-by-N lower triangular part of A contains the lower
00045 *          triangular part of the matrix A, and the strictly upper
00046 *          triangular part of A is not referenced.
00047 *
00048 *          On exit, if INFO = 0, the factor U or L from the Cholesky
00049 *          factorization A = U**T*U or A = L*L**T.
00050 *
00051 *  LDA     (input) INTEGER
00052 *          The leading dimension of the array A.  LDA >= max(1,N).
00053 *
00054 *  INFO    (output) INTEGER
00055 *          = 0:  successful exit
00056 *          < 0:  if INFO = -i, the i-th argument had an illegal value
00057 *          > 0:  if INFO = i, the leading minor of order i is not
00058 *                positive definite, and the factorization could not be
00059 *                completed.
00060 *
00061 *  =====================================================================
00062 *
00063 *     .. Parameters ..
00064       DOUBLE PRECISION   ONE
00065       PARAMETER          ( ONE = 1.0D+0 )
00066 *     ..
00067 *     .. Local Scalars ..
00068       LOGICAL            UPPER
00069       INTEGER            J, JB, NB
00070 *     ..
00071 *     .. External Functions ..
00072       LOGICAL            LSAME
00073       INTEGER            ILAENV
00074       EXTERNAL           LSAME, ILAENV
00075 *     ..
00076 *     .. External Subroutines ..
00077       EXTERNAL           DGEMM, DPOTF2, DSYRK, DTRSM, XERBLA
00078 *     ..
00079 *     .. Intrinsic Functions ..
00080       INTRINSIC          MAX, MIN
00081 *     ..
00082 *     .. Executable Statements ..
00083 *
00084 *     Test the input parameters.
00085 *
00086       INFO = 0
00087       UPPER = LSAME( UPLO, 'U' )
00088       IF( .NOT.UPPER .AND. .NOT.LSAME( UPLO, 'L' ) ) THEN
00089          INFO = -1
00090       ELSE IF( N.LT.0 ) THEN
00091          INFO = -2
00092       ELSE IF( LDA.LT.MAX( 1, N ) ) THEN
00093          INFO = -4
00094       END IF
00095       IF( INFO.NE.0 ) THEN
00096          CALL XERBLA( 'DPOTRF', -INFO )
00097          RETURN
00098       END IF
00099 *
00100 *     Quick return if possible
00101 *
00102       IF( N.EQ.0 )
00103      $   RETURN
00104 *
00105 *     Determine the block size for this environment.
00106 *
00107       NB = ILAENV( 1, 'DPOTRF', UPLO, N, -1, -1, -1 )
00108       IF( NB.LE.1 .OR. NB.GE.N ) THEN
00109 *
00110 *        Use unblocked code.
00111 *
00112          CALL DPOTF2( UPLO, N, A, LDA, INFO )
00113       ELSE
00114 *
00115 *        Use blocked code.
00116 *
00117          IF( UPPER ) THEN
00118 *
00119 *           Compute the Cholesky factorization A = U'*U.
00120 *
00121             DO 10 J = 1, N, NB
00122 *
00123 *              Update and factorize the current diagonal block and test
00124 *              for non-positive-definiteness.
00125 *
00126                JB = MIN( NB, N-J+1 )
00127                CALL DSYRK( 'Upper', 'Transpose', JB, J-1, -ONE,
00128      $                     A( 1, J ), LDA, ONE, A( J, J ), LDA )
00129                CALL DPOTF2( 'Upper', JB, A( J, J ), LDA, INFO )
00130                IF( INFO.NE.0 )
00131      $            GO TO 30
00132                IF( J+JB.LE.N ) THEN
00133 *
00134 *                 Compute the current block row.
00135 *
00136                   CALL DGEMM( 'Transpose', 'No transpose', JB, N-J-JB+1,
00137      $                        J-1, -ONE, A( 1, J ), LDA, A( 1, J+JB ),
00138      $                        LDA, ONE, A( J, J+JB ), LDA )
00139                   CALL DTRSM( 'Left', 'Upper', 'Transpose', 'Non-unit',
00140      $                        JB, N-J-JB+1, ONE, A( J, J ), LDA,
00141      $                        A( J, J+JB ), LDA )
00142                END IF
00143    10       CONTINUE
00144 *
00145          ELSE
00146 *
00147 *           Compute the Cholesky factorization A = L*L'.
00148 *
00149             DO 20 J = 1, N, NB
00150 *
00151 *              Update and factorize the current diagonal block and test
00152 *              for non-positive-definiteness.
00153 *
00154                JB = MIN( NB, N-J+1 )
00155                CALL DSYRK( 'Lower', 'No transpose', JB, J-1, -ONE,
00156      $                     A( J, 1 ), LDA, ONE, A( J, J ), LDA )
00157                CALL DPOTF2( 'Lower', JB, A( J, J ), LDA, INFO )
00158                IF( INFO.NE.0 )
00159      $            GO TO 30
00160                IF( J+JB.LE.N ) THEN
00161 *
00162 *                 Compute the current block column.
00163 *
00164                   CALL DGEMM( 'No transpose', 'Transpose', N-J-JB+1, JB,
00165      $                        J-1, -ONE, A( J+JB, 1 ), LDA, A( J, 1 ),
00166      $                        LDA, ONE, A( J+JB, J ), LDA )
00167                   CALL DTRSM( 'Right', 'Lower', 'Transpose', 'Non-unit',
00168      $                        N-J-JB+1, JB, ONE, A( J, J ), LDA,
00169      $                        A( J+JB, J ), LDA )
00170                END IF
00171    20       CONTINUE
00172          END IF
00173       END IF
00174       GO TO 40
00175 *
00176    30 CONTINUE
00177       INFO = INFO + J - 1
00178 *
00179    40 CONTINUE
00180       RETURN
00181 *
00182 *     End of DPOTRF
00183 *
00184       END