LAPACK: SRC/dopmtr.f Source File

Go to the documentation of this file.
00001       SUBROUTINE DOPMTR( SIDE, UPLO, TRANS, M, N, AP, TAU, C, LDC, WORK,
00002      $                   INFO )
00003 *
00004 *  -- LAPACK routine (version 3.2) --
00005 *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
00006 *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
00007 *     November 2006
00008 *
00009 *     .. Scalar Arguments ..
00010       CHARACTER          SIDE, TRANS, UPLO
00011       INTEGER            INFO, LDC, M, N
00012 *     ..
00013 *     .. Array Arguments ..
00014       DOUBLE PRECISION   AP( * ), C( LDC, * ), TAU( * ), WORK( * )
00015 *     ..
00016 *
00017 *  Purpose
00018 *  =======
00019 *
00020 *  DOPMTR overwrites the general real M-by-N matrix C with
00021 *
00022 *                  SIDE = 'L'     SIDE = 'R'
00023 *  TRANS = 'N':      Q * C          C * Q
00024 *  TRANS = 'T':      Q**T * C       C * Q**T
00025 *
00026 *  where Q is a real orthogonal matrix of order nq, with nq = m if
00027 *  SIDE = 'L' and nq = n if SIDE = 'R'. Q is defined as the product of
00028 *  nq-1 elementary reflectors, as returned by DSPTRD using packed
00029 *  storage:
00030 *
00031 *  if UPLO = 'U', Q = H(nq-1) . . . H(2) H(1);
00032 *
00033 *  if UPLO = 'L', Q = H(1) H(2) . . . H(nq-1).
00034 *
00035 *  Arguments
00036 *  =========
00037 *
00038 *  SIDE    (input) CHARACTER*1
00039 *          = 'L': apply Q or Q**T from the Left;
00040 *          = 'R': apply Q or Q**T from the Right.
00041 *
00042 *  UPLO    (input) CHARACTER*1
00043 *          = 'U': Upper triangular packed storage used in previous
00044 *                 call to DSPTRD;
00045 *          = 'L': Lower triangular packed storage used in previous
00046 *                 call to DSPTRD.
00047 *
00048 *  TRANS   (input) CHARACTER*1
00049 *          = 'N':  No transpose, apply Q;
00050 *          = 'T':  Transpose, apply Q**T.
00051 *
00052 *  M       (input) INTEGER
00053 *          The number of rows of the matrix C. M >= 0.
00054 *
00055 *  N       (input) INTEGER
00056 *          The number of columns of the matrix C. N >= 0.
00057 *
00058 *  AP      (input) DOUBLE PRECISION array, dimension
00059 *                               (M*(M+1)/2) if SIDE = 'L'
00060 *                               (N*(N+1)/2) if SIDE = 'R'
00061 *          The vectors which define the elementary reflectors, as
00062 *          returned by DSPTRD.  AP is modified by the routine but
00063 *          restored on exit.
00064 *
00065 *  TAU     (input) DOUBLE PRECISION array, dimension (M-1) if SIDE = 'L'
00066 *                                     or (N-1) if SIDE = 'R'
00067 *          TAU(i) must contain the scalar factor of the elementary
00068 *          reflector H(i), as returned by DSPTRD.
00069 *
00070 *  C       (input/output) DOUBLE PRECISION array, dimension (LDC,N)
00071 *          On entry, the M-by-N matrix C.
00072 *          On exit, C is overwritten by Q*C or Q**T*C or C*Q**T or C*Q.
00073 *
00074 *  LDC     (input) INTEGER
00075 *          The leading dimension of the array C. LDC >= max(1,M).
00076 *
00077 *  WORK    (workspace) DOUBLE PRECISION array, dimension
00078 *                                   (N) if SIDE = 'L'
00079 *                                   (M) if SIDE = 'R'
00080 *
00081 *  INFO    (output) INTEGER
00082 *          = 0:  successful exit
00083 *          < 0:  if INFO = -i, the i-th argument had an illegal value
00084 *
00085 *  =====================================================================
00086 *
00087 *     .. Parameters ..
00088       DOUBLE PRECISION   ONE
00089       PARAMETER          ( ONE = 1.0D+0 )
00090 *     ..
00091 *     .. Local Scalars ..
00092       LOGICAL            FORWRD, LEFT, NOTRAN, UPPER
00093       INTEGER            I, I1, I2, I3, IC, II, JC, MI, NI, NQ
00094       DOUBLE PRECISION   AII
00095 *     ..
00096 *     .. External Functions ..
00097       LOGICAL            LSAME
00098       EXTERNAL           LSAME
00099 *     ..
00100 *     .. External Subroutines ..
00101       EXTERNAL           DLARF, XERBLA
00102 *     ..
00103 *     .. Intrinsic Functions ..
00104       INTRINSIC          MAX
00105 *     ..
00106 *     .. Executable Statements ..
00107 *
00108 *     Test the input arguments
00109 *
00110       INFO = 0
00111       LEFT = LSAME( SIDE, 'L' )
00112       NOTRAN = LSAME( TRANS, 'N' )
00113       UPPER = LSAME( UPLO, 'U' )
00114 *
00115 *     NQ is the order of Q
00116 *
00117       IF( LEFT ) THEN
00118          NQ = M
00119       ELSE
00120          NQ = N
00121       END IF
00122       IF( .NOT.LEFT .AND. .NOT.LSAME( SIDE, 'R' ) ) THEN
00123          INFO = -1
00124       ELSE IF( .NOT.UPPER .AND. .NOT.LSAME( UPLO, 'L' ) ) THEN
00125          INFO = -2
00126       ELSE IF( .NOT.NOTRAN .AND. .NOT.LSAME( TRANS, 'T' ) ) THEN
00127          INFO = -3
00128       ELSE IF( M.LT.0 ) THEN
00129          INFO = -4
00130       ELSE IF( N.LT.0 ) THEN
00131          INFO = -5
00132       ELSE IF( LDC.LT.MAX( 1, M ) ) THEN
00133          INFO = -9
00134       END IF
00135       IF( INFO.NE.0 ) THEN
00136          CALL XERBLA( 'DOPMTR', -INFO )
00137          RETURN
00138       END IF
00139 *
00140 *     Quick return if possible
00141 *
00142       IF( M.EQ.0 .OR. N.EQ.0 )
00143      $   RETURN
00144 *
00145       IF( UPPER ) THEN
00146 *
00147 *        Q was determined by a call to DSPTRD with UPLO = 'U'
00148 *
00149          FORWRD = ( LEFT .AND. NOTRAN ) .OR.
00150      $            ( .NOT.LEFT .AND. .NOT.NOTRAN )
00151 *
00152          IF( FORWRD ) THEN
00153             I1 = 1
00154             I2 = NQ - 1
00155             I3 = 1
00156             II = 2
00157          ELSE
00158             I1 = NQ - 1
00159             I2 = 1
00160             I3 = -1
00161             II = NQ*( NQ+1 ) / 2 - 1
00162          END IF
00163 *
00164          IF( LEFT ) THEN
00165             NI = N
00166          ELSE
00167             MI = M
00168          END IF
00169 *
00170          DO 10 I = I1, I2, I3
00171             IF( LEFT ) THEN
00172 *
00173 *              H(i) is applied to C(1:i,1:n)
00174 *
00175                MI = I
00176             ELSE
00177 *
00178 *              H(i) is applied to C(1:m,1:i)
00179 *
00180                NI = I
00181             END IF
00182 *
00183 *           Apply H(i)
00184 *
00185             AII = AP( II )
00186             AP( II ) = ONE
00187             CALL DLARF( SIDE, MI, NI, AP( II-I+1 ), 1, TAU( I ), C, LDC,
00188      $                  WORK )
00189             AP( II ) = AII
00190 *
00191             IF( FORWRD ) THEN
00192                II = II + I + 2
00193             ELSE
00194                II = II - I - 1
00195             END IF
00196    10    CONTINUE
00197       ELSE
00198 *
00199 *        Q was determined by a call to DSPTRD with UPLO = 'L'.
00200 *
00201          FORWRD = ( LEFT .AND. .NOT.NOTRAN ) .OR.
00202      $            ( .NOT.LEFT .AND. NOTRAN )
00203 *
00204          IF( FORWRD ) THEN
00205             I1 = 1
00206             I2 = NQ - 1
00207             I3 = 1
00208             II = 2
00209          ELSE
00210             I1 = NQ - 1
00211             I2 = 1
00212             I3 = -1
00213             II = NQ*( NQ+1 ) / 2 - 1
00214          END IF
00215 *
00216          IF( LEFT ) THEN
00217             NI = N
00218             JC = 1
00219          ELSE
00220             MI = M
00221             IC = 1
00222          END IF
00223 *
00224          DO 20 I = I1, I2, I3
00225             AII = AP( II )
00226             AP( II ) = ONE
00227             IF( LEFT ) THEN
00228 *
00229 *              H(i) is applied to C(i+1:m,1:n)
00230 *
00231                MI = M - I
00232                IC = I + 1
00233             ELSE
00234 *
00235 *              H(i) is applied to C(1:m,i+1:n)
00236 *
00237                NI = N - I
00238                JC = I + 1
00239             END IF
00240 *
00241 *           Apply H(i)
00242 *
00243             CALL DLARF( SIDE, MI, NI, AP( II ), 1, TAU( I ),
00244      $                  C( IC, JC ), LDC, WORK )
00245             AP( II ) = AII
00246 *
00247             IF( FORWRD ) THEN
00248                II = II + NQ - I + 1
00249             ELSE
00250                II = II - NQ + I - 2
00251             END IF
00252    20    CONTINUE
00253       END IF
00254       RETURN
00255 *
00256 *     End of DOPMTR
00257 *
00258       END