Statistiques
| Révision :

root / src / lapack / double / dlarzt.f @ 10

Historique | Voir | Annoter | Télécharger (5,87 ko)

1
      SUBROUTINE DLARZT( DIRECT, STOREV, N, K, V, LDV, TAU, T, LDT )
2
*
3
*  -- LAPACK routine (version 3.2) --
4
*  -- LAPACK is a software package provided by Univ. of Tennessee,    --
5
*  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
6
*     November 2006
7
*
8
*     .. Scalar Arguments ..
9
      CHARACTER          DIRECT, STOREV
10
      INTEGER            K, LDT, LDV, N
11
*     ..
12
*     .. Array Arguments ..
13
      DOUBLE PRECISION   T( LDT, * ), TAU( * ), V( LDV, * )
14
*     ..
15
*
16
*  Purpose
17
*  =======
18
*
19
*  DLARZT forms the triangular factor T of a real block reflector
20
*  H of order > n, which is defined as a product of k elementary
21
*  reflectors.
22
*
23
*  If DIRECT = 'F', H = H(1) H(2) . . . H(k) and T is upper triangular;
24
*
25
*  If DIRECT = 'B', H = H(k) . . . H(2) H(1) and T is lower triangular.
26
*
27
*  If STOREV = 'C', the vector which defines the elementary reflector
28
*  H(i) is stored in the i-th column of the array V, and
29
*
30
*     H  =  I - V * T * V'
31
*
32
*  If STOREV = 'R', the vector which defines the elementary reflector
33
*  H(i) is stored in the i-th row of the array V, and
34
*
35
*     H  =  I - V' * T * V
36
*
37
*  Currently, only STOREV = 'R' and DIRECT = 'B' are supported.
38
*
39
*  Arguments
40
*  =========
41
*
42
*  DIRECT  (input) CHARACTER*1
43
*          Specifies the order in which the elementary reflectors are
44
*          multiplied to form the block reflector:
45
*          = 'F': H = H(1) H(2) . . . H(k) (Forward, not supported yet)
46
*          = 'B': H = H(k) . . . H(2) H(1) (Backward)
47
*
48
*  STOREV  (input) CHARACTER*1
49
*          Specifies how the vectors which define the elementary
50
*          reflectors are stored (see also Further Details):
51
*          = 'C': columnwise                        (not supported yet)
52
*          = 'R': rowwise
53
*
54
*  N       (input) INTEGER
55
*          The order of the block reflector H. N >= 0.
56
*
57
*  K       (input) INTEGER
58
*          The order of the triangular factor T (= the number of
59
*          elementary reflectors). K >= 1.
60
*
61
*  V       (input/output) DOUBLE PRECISION array, dimension
62
*                               (LDV,K) if STOREV = 'C'
63
*                               (LDV,N) if STOREV = 'R'
64
*          The matrix V. See further details.
65
*
66
*  LDV     (input) INTEGER
67
*          The leading dimension of the array V.
68
*          If STOREV = 'C', LDV >= max(1,N); if STOREV = 'R', LDV >= K.
69
*
70
*  TAU     (input) DOUBLE PRECISION array, dimension (K)
71
*          TAU(i) must contain the scalar factor of the elementary
72
*          reflector H(i).
73
*
74
*  T       (output) DOUBLE PRECISION array, dimension (LDT,K)
75
*          The k by k triangular factor T of the block reflector.
76
*          If DIRECT = 'F', T is upper triangular; if DIRECT = 'B', T is
77
*          lower triangular. The rest of the array is not used.
78
*
79
*  LDT     (input) INTEGER
80
*          The leading dimension of the array T. LDT >= K.
81
*
82
*  Further Details
83
*  ===============
84
*
85
*  Based on contributions by
86
*    A. Petitet, Computer Science Dept., Univ. of Tenn., Knoxville, USA
87
*
88
*  The shape of the matrix V and the storage of the vectors which define
89
*  the H(i) is best illustrated by the following example with n = 5 and
90
*  k = 3. The elements equal to 1 are not stored; the corresponding
91
*  array elements are modified but restored on exit. The rest of the
92
*  array is not used.
93
*
94
*  DIRECT = 'F' and STOREV = 'C':         DIRECT = 'F' and STOREV = 'R':
95
*
96
*                                              ______V_____
97
*         ( v1 v2 v3 )                        /            \
98
*         ( v1 v2 v3 )                      ( v1 v1 v1 v1 v1 . . . . 1 )
99
*     V = ( v1 v2 v3 )                      ( v2 v2 v2 v2 v2 . . . 1   )
100
*         ( v1 v2 v3 )                      ( v3 v3 v3 v3 v3 . . 1     )
101
*         ( v1 v2 v3 )
102
*            .  .  .
103
*            .  .  .
104
*            1  .  .
105
*               1  .
106
*                  1
107
*
108
*  DIRECT = 'B' and STOREV = 'C':         DIRECT = 'B' and STOREV = 'R':
109
*
110
*                                                        ______V_____
111
*            1                                          /            \
112
*            .  1                           ( 1 . . . . v1 v1 v1 v1 v1 )
113
*            .  .  1                        ( . 1 . . . v2 v2 v2 v2 v2 )
114
*            .  .  .                        ( . . 1 . . v3 v3 v3 v3 v3 )
115
*            .  .  .
116
*         ( v1 v2 v3 )
117
*         ( v1 v2 v3 )
118
*     V = ( v1 v2 v3 )
119
*         ( v1 v2 v3 )
120
*         ( v1 v2 v3 )
121
*
122
*  =====================================================================
123
*
124
*     .. Parameters ..
125
      DOUBLE PRECISION   ZERO
126
      PARAMETER          ( ZERO = 0.0D+0 )
127
*     ..
128
*     .. Local Scalars ..
129
      INTEGER            I, INFO, J
130
*     ..
131
*     .. External Subroutines ..
132
      EXTERNAL           DGEMV, DTRMV, XERBLA
133
*     ..
134
*     .. External Functions ..
135
      LOGICAL            LSAME
136
      EXTERNAL           LSAME
137
*     ..
138
*     .. Executable Statements ..
139
*
140
*     Check for currently supported options
141
*
142
      INFO = 0
143
      IF( .NOT.LSAME( DIRECT, 'B' ) ) THEN
144
         INFO = -1
145
      ELSE IF( .NOT.LSAME( STOREV, 'R' ) ) THEN
146
         INFO = -2
147
      END IF
148
      IF( INFO.NE.0 ) THEN
149
         CALL XERBLA( 'DLARZT', -INFO )
150
         RETURN
151
      END IF
152
*
153
      DO 20 I = K, 1, -1
154
         IF( TAU( I ).EQ.ZERO ) THEN
155
*
156
*           H(i)  =  I
157
*
158
            DO 10 J = I, K
159
               T( J, I ) = ZERO
160
   10       CONTINUE
161
         ELSE
162
*
163
*           general case
164
*
165
            IF( I.LT.K ) THEN
166
*
167
*              T(i+1:k,i) = - tau(i) * V(i+1:k,1:n) * V(i,1:n)'
168
*
169
               CALL DGEMV( 'No transpose', K-I, N, -TAU( I ),
170
     $                     V( I+1, 1 ), LDV, V( I, 1 ), LDV, ZERO,
171
     $                     T( I+1, I ), 1 )
172
*
173
*              T(i+1:k,i) = T(i+1:k,i+1:k) * T(i+1:k,i)
174
*
175
               CALL DTRMV( 'Lower', 'No transpose', 'Non-unit', K-I,
176
     $                     T( I+1, I+1 ), LDT, T( I+1, I ), 1 )
177
            END IF
178
            T( I, I ) = TAU( I )
179
         END IF
180
   20 CONTINUE
181
      RETURN
182
*
183
*     End of DLARZT
184
*
185
      END