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
|
#include "rb_lapack.h"
extern VOID sgesdd_(char* jobz, integer* m, integer* n, real* a, integer* lda, real* s, real* u, integer* ldu, real* vt, integer* ldvt, real* work, integer* lwork, integer* iwork, integer* info);
static VALUE
rblapack_sgesdd(int argc, VALUE *argv, VALUE self){
VALUE rblapack_jobz;
char jobz;
VALUE rblapack_a;
real *a;
VALUE rblapack_lwork;
integer lwork;
VALUE rblapack_s;
real *s;
VALUE rblapack_u;
real *u;
VALUE rblapack_vt;
real *vt;
VALUE rblapack_work;
real *work;
VALUE rblapack_info;
integer info;
VALUE rblapack_a_out__;
real *a_out__;
integer *iwork;
integer lda;
integer n;
integer ldu;
integer ucol;
integer ldvt;
integer m;
VALUE rblapack_options;
if (argc > 0 && TYPE(argv[argc-1]) == T_HASH) {
argc--;
rblapack_options = argv[argc];
if (rb_hash_aref(rblapack_options, sHelp) == Qtrue) {
printf("%s\n", "USAGE:\n s, u, vt, work, info, a = NumRu::Lapack.sgesdd( jobz, a, [:lwork => lwork, :usage => usage, :help => help])\n\n\nFORTRAN MANUAL\n SUBROUTINE SGESDD( JOBZ, M, N, A, LDA, S, U, LDU, VT, LDVT, WORK, LWORK, IWORK, INFO )\n\n* Purpose\n* =======\n*\n* SGESDD computes the singular value decomposition (SVD) of a real\n* M-by-N matrix A, optionally computing the left and right singular\n* vectors. If singular vectors are desired, it uses a\n* divide-and-conquer algorithm.\n*\n* The SVD is written\n*\n* A = U * SIGMA * transpose(V)\n*\n* where SIGMA is an M-by-N matrix which is zero except for its\n* min(m,n) diagonal elements, U is an M-by-M orthogonal matrix, and\n* V is an N-by-N orthogonal matrix. The diagonal elements of SIGMA\n* are the singular values of A; they are real and non-negative, and\n* are returned in descending order. The first min(m,n) columns of\n* U and V are the left and right singular vectors of A.\n*\n* Note that the routine returns VT = V**T, not V.\n*\n* The divide and conquer algorithm makes very mild assumptions about\n* floating point arithmetic. It will work on machines with a guard\n* digit in add/subtract, or on those binary machines without guard\n* digits which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or\n* Cray-2. It could conceivably fail on hexadecimal or decimal machines\n* without guard digits, but we know of none.\n*\n\n* Arguments\n* =========\n*\n* JOBZ (input) CHARACTER*1\n* Specifies options for computing all or part of the matrix U:\n* = 'A': all M columns of U and all N rows of V**T are\n* returned in the arrays U and VT;\n* = 'S': the first min(M,N) columns of U and the first\n* min(M,N) rows of V**T are returned in the arrays U\n* and VT;\n* = 'O': If M >= N, the first N columns of U are overwritten\n* on the array A and all rows of V**T are returned in\n* the array VT;\n* otherwise, all columns of U are returned in the\n* array U and the first M rows of V**T are overwritten\n* in the array A;\n* = 'N': no columns of U or rows of V**T are computed.\n*\n* M (input) INTEGER\n* The number of rows of the input matrix A. M >= 0.\n*\n* N (input) INTEGER\n* The number of columns of the input matrix A. N >= 0.\n*\n* A (input/output) REAL array, dimension (LDA,N)\n* On entry, the M-by-N matrix A.\n* On exit,\n* if JOBZ = 'O', A is overwritten with the first N columns\n* of U (the left singular vectors, stored\n* columnwise) if M >= N;\n* A is overwritten with the first M rows\n* of V**T (the right singular vectors, stored\n* rowwise) otherwise.\n* if JOBZ .ne. 'O', the contents of A are destroyed.\n*\n* LDA (input) INTEGER\n* The leading dimension of the array A. LDA >= max(1,M).\n*\n* S (output) REAL array, dimension (min(M,N))\n* The singular values of A, sorted so that S(i) >= S(i+1).\n*\n* U (output) REAL array, dimension (LDU,UCOL)\n* UCOL = M if JOBZ = 'A' or JOBZ = 'O' and M < N;\n* UCOL = min(M,N) if JOBZ = 'S'.\n* If JOBZ = 'A' or JOBZ = 'O' and M < N, U contains the M-by-M\n* orthogonal matrix U;\n* if JOBZ = 'S', U contains the first min(M,N) columns of U\n* (the left singular vectors, stored columnwise);\n* if JOBZ = 'O' and M >= N, or JOBZ = 'N', U is not referenced.\n*\n* LDU (input) INTEGER\n* The leading dimension of the array U. LDU >= 1; if\n* JOBZ = 'S' or 'A' or JOBZ = 'O' and M < N, LDU >= M.\n*\n* VT (output) REAL array, dimension (LDVT,N)\n* If JOBZ = 'A' or JOBZ = 'O' and M >= N, VT contains the\n* N-by-N orthogonal matrix V**T;\n* if JOBZ = 'S', VT contains the first min(M,N) rows of\n* V**T (the right singular vectors, stored rowwise);\n* if JOBZ = 'O' and M < N, or JOBZ = 'N', VT is not referenced.\n*\n* LDVT (input) INTEGER\n* The leading dimension of the array VT. LDVT >= 1; if\n* JOBZ = 'A' or JOBZ = 'O' and M >= N, LDVT >= N;\n* if JOBZ = 'S', LDVT >= min(M,N).\n*\n* WORK (workspace/output) REAL array, dimension (MAX(1,LWORK))\n* On exit, if INFO = 0, WORK(1) returns the optimal LWORK;\n*\n* LWORK (input) INTEGER\n* The dimension of the array WORK. LWORK >= 1.\n* If JOBZ = 'N',\n* LWORK >= 3*min(M,N) + max(max(M,N),6*min(M,N)).\n* If JOBZ = 'O',\n* LWORK >= 3*min(M,N) + \n* max(max(M,N),5*min(M,N)*min(M,N)+4*min(M,N)).\n* If JOBZ = 'S' or 'A'\n* LWORK >= 3*min(M,N) +\n* max(max(M,N),4*min(M,N)*min(M,N)+4*min(M,N)).\n* For good performance, LWORK should generally be larger.\n* If LWORK = -1 but other input arguments are legal, WORK(1)\n* returns the optimal LWORK.\n*\n* IWORK (workspace) INTEGER array, dimension (8*min(M,N))\n*\n* INFO (output) INTEGER\n* = 0: successful exit.\n* < 0: if INFO = -i, the i-th argument had an illegal value.\n* > 0: SBDSDC did not converge, updating process failed.\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\n");
return Qnil;
}
if (rb_hash_aref(rblapack_options, sUsage) == Qtrue) {
printf("%s\n", "USAGE:\n s, u, vt, work, info, a = NumRu::Lapack.sgesdd( jobz, a, [:lwork => lwork, :usage => usage, :help => help])\n");
return Qnil;
}
} else
rblapack_options = Qnil;
if (argc != 2 && argc != 3)
rb_raise(rb_eArgError,"wrong number of arguments (%d for 2)", argc);
rblapack_jobz = argv[0];
rblapack_a = argv[1];
if (argc == 3) {
rblapack_lwork = argv[2];
} else if (rblapack_options != Qnil) {
rblapack_lwork = rb_hash_aref(rblapack_options, ID2SYM(rb_intern("lwork")));
} else {
rblapack_lwork = Qnil;
}
jobz = StringValueCStr(rblapack_jobz)[0];
if (!NA_IsNArray(rblapack_a))
rb_raise(rb_eArgError, "a (2th argument) must be NArray");
if (NA_RANK(rblapack_a) != 2)
rb_raise(rb_eArgError, "rank of a (2th argument) must be %d", 2);
lda = NA_SHAPE0(rblapack_a);
n = NA_SHAPE1(rblapack_a);
if (NA_TYPE(rblapack_a) != NA_SFLOAT)
rblapack_a = na_change_type(rblapack_a, NA_SFLOAT);
a = NA_PTR_TYPE(rblapack_a, real*);
m = lda;
ldvt = ((lsame_(&jobz,"A")) || (((lsame_(&jobz,"O")) && (m >= n)))) ? n : lsame_(&jobz,"S") ? MIN(m,n) : 1;
if (rblapack_lwork == Qnil)
lwork = lsame_(&jobz,"N") ? 3*MIN(m,n)+MAX(MAX(m,n),7*MIN(m,n)) : lsame_(&jobz,"O") ? 3*MIN(m,n)+MAX(MAX(m,n),5*MIN(m,n)*MIN(m,n)+4*MIN(m,n)) : (lsame_(&jobz,"S")||lsame_(&jobz,"A")) ? 3*MIN(m,n)+MAX(MAX(m,n),4*MIN(m,n)*MIN(m,n)+4*MIN(m,n)) : 0;
else {
lwork = NUM2INT(rblapack_lwork);
}
ldu = (lsame_(&jobz,"S") || lsame_(&jobz,"A") || (lsame_(&jobz,"O") && m < n)) ? m : 1;
ucol = ((lsame_(&jobz,"A")) || (((lsame_(&jobz,"O")) && (m < n)))) ? m : lsame_(&jobz,"S") ? MIN(m,n) : 0;
{
na_shape_t shape[1];
shape[0] = MIN(m,n);
rblapack_s = na_make_object(NA_SFLOAT, 1, shape, cNArray);
}
s = NA_PTR_TYPE(rblapack_s, real*);
{
na_shape_t shape[2];
shape[0] = ldu;
shape[1] = ucol;
rblapack_u = na_make_object(NA_SFLOAT, 2, shape, cNArray);
}
u = NA_PTR_TYPE(rblapack_u, real*);
{
na_shape_t shape[2];
shape[0] = ldvt;
shape[1] = n;
rblapack_vt = na_make_object(NA_SFLOAT, 2, shape, cNArray);
}
vt = NA_PTR_TYPE(rblapack_vt, real*);
{
na_shape_t shape[1];
shape[0] = MAX(1,lwork);
rblapack_work = na_make_object(NA_SFLOAT, 1, shape, cNArray);
}
work = NA_PTR_TYPE(rblapack_work, real*);
{
na_shape_t shape[2];
shape[0] = lda;
shape[1] = n;
rblapack_a_out__ = na_make_object(NA_SFLOAT, 2, shape, cNArray);
}
a_out__ = NA_PTR_TYPE(rblapack_a_out__, real*);
MEMCPY(a_out__, a, real, NA_TOTAL(rblapack_a));
rblapack_a = rblapack_a_out__;
a = a_out__;
iwork = ALLOC_N(integer, (8*MIN(m,n)));
sgesdd_(&jobz, &m, &n, a, &lda, s, u, &ldu, vt, &ldvt, work, &lwork, iwork, &info);
free(iwork);
rblapack_info = INT2NUM(info);
return rb_ary_new3(6, rblapack_s, rblapack_u, rblapack_vt, rblapack_work, rblapack_info, rblapack_a);
}
void
init_lapack_sgesdd(VALUE mLapack, VALUE sH, VALUE sU, VALUE zero){
sHelp = sH;
sUsage = sU;
rblapack_ZERO = zero;
rb_define_module_function(mLapack, "sgesdd", rblapack_sgesdd, -1);
}
|
