1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
|
---
:name: dlasd0
:md5sum: 469c67fd3b988986d8c9fe190b5a3db7
:category: :subroutine
:arguments:
- n:
:type: integer
:intent: input
- sqre:
:type: integer
:intent: input
- d:
:type: doublereal
:intent: input/output
:dims:
- n
- e:
:type: doublereal
:intent: input
:dims:
- m-1
- u:
:type: doublereal
:intent: output
:dims:
- ldu
- n
- ldu:
:type: integer
:intent: input
- vt:
:type: doublereal
:intent: output
:dims:
- ldvt
- m
- ldvt:
:type: integer
:intent: input
- smlsiz:
:type: integer
:intent: input
- iwork:
:type: integer
:intent: workspace
:dims:
- (8 * n)
- work:
:type: doublereal
:intent: workspace
:dims:
- (3 * pow(m,2) + 2 * m)
- info:
:type: integer
:intent: output
:substitutions:
m: "sqre == 0 ? n : sqre == 1 ? n+1 : 0"
ldvt: n
ldu: n
:fortran_help: " SUBROUTINE DLASD0( N, SQRE, D, E, U, LDU, VT, LDVT, SMLSIZ, IWORK, WORK, INFO )\n\n\
* Purpose\n\
* =======\n\
*\n\
* Using a divide and conquer approach, DLASD0 computes the singular\n\
* value decomposition (SVD) of a real upper bidiagonal N-by-M\n\
* matrix B with diagonal D and offdiagonal E, where M = N + SQRE.\n\
* The algorithm computes orthogonal matrices U and VT such that\n\
* B = U * S * VT. The singular values S are overwritten on D.\n\
*\n\
* A related subroutine, DLASDA, computes only the singular values,\n\
* and optionally, the singular vectors in compact form.\n\
*\n\n\
* Arguments\n\
* =========\n\
*\n\
* N (input) INTEGER\n\
* On entry, the row dimension of the upper bidiagonal matrix.\n\
* This is also the dimension of the main diagonal array D.\n\
*\n\
* SQRE (input) INTEGER\n\
* Specifies the column dimension of the bidiagonal matrix.\n\
* = 0: The bidiagonal matrix has column dimension M = N;\n\
* = 1: The bidiagonal matrix has column dimension M = N+1;\n\
*\n\
* D (input/output) DOUBLE PRECISION array, dimension (N)\n\
* On entry D contains the main diagonal of the bidiagonal\n\
* matrix.\n\
* On exit D, if INFO = 0, contains its singular values.\n\
*\n\
* E (input) DOUBLE PRECISION array, dimension (M-1)\n\
* Contains the subdiagonal entries of the bidiagonal matrix.\n\
* On exit, E has been destroyed.\n\
*\n\
* U (output) DOUBLE PRECISION array, dimension at least (LDQ, N)\n\
* On exit, U contains the left singular vectors.\n\
*\n\
* LDU (input) INTEGER\n\
* On entry, leading dimension of U.\n\
*\n\
* VT (output) DOUBLE PRECISION array, dimension at least (LDVT, M)\n\
* On exit, VT' contains the right singular vectors.\n\
*\n\
* LDVT (input) INTEGER\n\
* On entry, leading dimension of VT.\n\
*\n\
* SMLSIZ (input) INTEGER\n\
* On entry, maximum size of the subproblems at the\n\
* bottom of the computation tree.\n\
*\n\
* IWORK (workspace) INTEGER work array.\n\
* Dimension must be at least (8 * N)\n\
*\n\
* WORK (workspace) DOUBLE PRECISION work array.\n\
* Dimension must be at least (3 * M**2 + 2 * M)\n\
*\n\
* INFO (output) INTEGER\n\
* = 0: successful exit.\n\
* < 0: if INFO = -i, the i-th argument had an illegal value.\n\
* > 0: if INFO = 1, a singular value did not converge\n\
*\n\n\
* Further Details\n\
* ===============\n\
*\n\
* Based on contributions by\n\
* Ming Gu and Huan Ren, Computer Science Division, University of\n\
* California at Berkeley, USA\n\
*\n\
* =====================================================================\n\
*\n\
* .. Local Scalars ..\n INTEGER I, I1, IC, IDXQ, IDXQC, IM1, INODE, ITEMP, IWK,\n $ J, LF, LL, LVL, M, NCC, ND, NDB1, NDIML, NDIMR,\n $ NL, NLF, NLP1, NLVL, NR, NRF, NRP1, SQREI\n DOUBLE PRECISION ALPHA, BETA\n\
* ..\n\
* .. External Subroutines ..\n EXTERNAL DLASD1, DLASDQ, DLASDT, XERBLA\n\
* ..\n"
|
