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
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
|
SUBROUTINE SGGGLM( N, M, P, A, LDA, B, LDB, D, X, Y, WORK, LWORK,
$ INFO )
*
* -- LAPACK driver routine (version 3.3.1) --
* -- LAPACK is a software package provided by Univ. of Tennessee, --
* -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
* -- April 2011 --
*
* .. Scalar Arguments ..
INTEGER INFO, LDA, LDB, LWORK, M, N, P
* ..
* .. Array Arguments ..
REAL A( LDA, * ), B( LDB, * ), D( * ), WORK( * ),
$ X( * ), Y( * )
* ..
*
* Purpose
* =======
*
* SGGGLM solves a general Gauss-Markov linear model (GLM) problem:
*
* minimize || y ||_2 subject to d = A*x + B*y
* x
*
* where A is an N-by-M matrix, B is an N-by-P matrix, and d is a
* given N-vector. It is assumed that M <= N <= M+P, and
*
* rank(A) = M and rank( A B ) = N.
*
* Under these assumptions, the constrained equation is always
* consistent, and there is a unique solution x and a minimal 2-norm
* solution y, which is obtained using a generalized QR factorization
* of the matrices (A, B) given by
*
* A = Q*(R), B = Q*T*Z.
* (0)
*
* In particular, if matrix B is square nonsingular, then the problem
* GLM is equivalent to the following weighted linear least squares
* problem
*
* minimize || inv(B)*(d-A*x) ||_2
* x
*
* where inv(B) denotes the inverse of B.
*
* Arguments
* =========
*
* N (input) INTEGER
* The number of rows of the matrices A and B. N >= 0.
*
* M (input) INTEGER
* The number of columns of the matrix A. 0 <= M <= N.
*
* P (input) INTEGER
* The number of columns of the matrix B. P >= N-M.
*
* A (input/output) REAL array, dimension (LDA,M)
* On entry, the N-by-M matrix A.
* On exit, the upper triangular part of the array A contains
* the M-by-M upper triangular matrix R.
*
* LDA (input) INTEGER
* The leading dimension of the array A. LDA >= max(1,N).
*
* B (input/output) REAL array, dimension (LDB,P)
* On entry, the N-by-P matrix B.
* On exit, if N <= P, the upper triangle of the subarray
* B(1:N,P-N+1:P) contains the N-by-N upper triangular matrix T;
* if N > P, the elements on and above the (N-P)th subdiagonal
* contain the N-by-P upper trapezoidal matrix T.
*
* LDB (input) INTEGER
* The leading dimension of the array B. LDB >= max(1,N).
*
* D (input/output) REAL array, dimension (N)
* On entry, D is the left hand side of the GLM equation.
* On exit, D is destroyed.
*
* X (output) REAL array, dimension (M)
* Y (output) REAL array, dimension (P)
* On exit, X and Y are the solutions of the GLM problem.
*
* WORK (workspace/output) REAL array, dimension (MAX(1,LWORK))
* On exit, if INFO = 0, WORK(1) returns the optimal LWORK.
*
* LWORK (input) INTEGER
* The dimension of the array WORK. LWORK >= max(1,N+M+P).
* For optimum performance, LWORK >= M+min(N,P)+max(N,P)*NB,
* where NB is an upper bound for the optimal blocksizes for
* SGEQRF, SGERQF, SORMQR and SORMRQ.
*
* If LWORK = -1, then a workspace query is assumed; the routine
* only calculates the optimal size of the WORK array, returns
* this value as the first entry of the WORK array, and no error
* message related to LWORK is issued by XERBLA.
*
* INFO (output) INTEGER
* = 0: successful exit.
* < 0: if INFO = -i, the i-th argument had an illegal value.
* = 1: the upper triangular factor R associated with A in the
* generalized QR factorization of the pair (A, B) is
* singular, so that rank(A) < M; the least squares
* solution could not be computed.
* = 2: the bottom (N-M) by (N-M) part of the upper trapezoidal
* factor T associated with B in the generalized QR
* factorization of the pair (A, B) is singular, so that
* rank( A B ) < N; the least squares solution could not
* be computed.
*
* ===================================================================
*
* .. Parameters ..
REAL ZERO, ONE
PARAMETER ( ZERO = 0.0E+0, ONE = 1.0E+0 )
* ..
* .. Local Scalars ..
LOGICAL LQUERY
INTEGER I, LOPT, LWKMIN, LWKOPT, NB, NB1, NB2, NB3,
$ NB4, NP
* ..
* .. External Subroutines ..
EXTERNAL SCOPY, SGEMV, SGGQRF, SORMQR, SORMRQ, STRTRS,
$ XERBLA
* ..
* .. External Functions ..
INTEGER ILAENV
EXTERNAL ILAENV
* ..
* .. Intrinsic Functions ..
INTRINSIC INT, MAX, MIN
* ..
* .. Executable Statements ..
*
* Test the input parameters
*
INFO = 0
NP = MIN( N, P )
LQUERY = ( LWORK.EQ.-1 )
IF( N.LT.0 ) THEN
INFO = -1
ELSE IF( M.LT.0 .OR. M.GT.N ) THEN
INFO = -2
ELSE IF( P.LT.0 .OR. P.LT.N-M ) THEN
INFO = -3
ELSE IF( LDA.LT.MAX( 1, N ) ) THEN
INFO = -5
ELSE IF( LDB.LT.MAX( 1, N ) ) THEN
INFO = -7
END IF
*
* Calculate workspace
*
IF( INFO.EQ.0) THEN
IF( N.EQ.0 ) THEN
LWKMIN = 1
LWKOPT = 1
ELSE
NB1 = ILAENV( 1, 'SGEQRF', ' ', N, M, -1, -1 )
NB2 = ILAENV( 1, 'SGERQF', ' ', N, M, -1, -1 )
NB3 = ILAENV( 1, 'SORMQR', ' ', N, M, P, -1 )
NB4 = ILAENV( 1, 'SORMRQ', ' ', N, M, P, -1 )
NB = MAX( NB1, NB2, NB3, NB4 )
LWKMIN = M + N + P
LWKOPT = M + NP + MAX( N, P )*NB
END IF
WORK( 1 ) = LWKOPT
*
IF( LWORK.LT.LWKMIN .AND. .NOT.LQUERY ) THEN
INFO = -12
END IF
END IF
*
IF( INFO.NE.0 ) THEN
CALL XERBLA( 'SGGGLM', -INFO )
RETURN
ELSE IF( LQUERY ) THEN
RETURN
END IF
*
* Quick return if possible
*
IF( N.EQ.0 )
$ RETURN
*
* Compute the GQR factorization of matrices A and B:
*
* Q**T*A = ( R11 ) M, Q**T*B*Z**T = ( T11 T12 ) M
* ( 0 ) N-M ( 0 T22 ) N-M
* M M+P-N N-M
*
* where R11 and T22 are upper triangular, and Q and Z are
* orthogonal.
*
CALL SGGQRF( N, M, P, A, LDA, WORK, B, LDB, WORK( M+1 ),
$ WORK( M+NP+1 ), LWORK-M-NP, INFO )
LOPT = WORK( M+NP+1 )
*
* Update left-hand-side vector d = Q**T*d = ( d1 ) M
* ( d2 ) N-M
*
CALL SORMQR( 'Left', 'Transpose', N, 1, M, A, LDA, WORK, D,
$ MAX( 1, N ), WORK( M+NP+1 ), LWORK-M-NP, INFO )
LOPT = MAX( LOPT, INT( WORK( M+NP+1 ) ) )
*
* Solve T22*y2 = d2 for y2
*
IF( N.GT.M ) THEN
CALL STRTRS( 'Upper', 'No transpose', 'Non unit', N-M, 1,
$ B( M+1, M+P-N+1 ), LDB, D( M+1 ), N-M, INFO )
*
IF( INFO.GT.0 ) THEN
INFO = 1
RETURN
END IF
*
CALL SCOPY( N-M, D( M+1 ), 1, Y( M+P-N+1 ), 1 )
END IF
*
* Set y1 = 0
*
DO 10 I = 1, M + P - N
Y( I ) = ZERO
10 CONTINUE
*
* Update d1 = d1 - T12*y2
*
CALL SGEMV( 'No transpose', M, N-M, -ONE, B( 1, M+P-N+1 ), LDB,
$ Y( M+P-N+1 ), 1, ONE, D, 1 )
*
* Solve triangular system: R11*x = d1
*
IF( M.GT.0 ) THEN
CALL STRTRS( 'Upper', 'No Transpose', 'Non unit', M, 1, A, LDA,
$ D, M, INFO )
*
IF( INFO.GT.0 ) THEN
INFO = 2
RETURN
END IF
*
* Copy D to X
*
CALL SCOPY( M, D, 1, X, 1 )
END IF
*
* Backward transformation y = Z**T *y
*
CALL SORMRQ( 'Left', 'Transpose', P, 1, NP,
$ B( MAX( 1, N-P+1 ), 1 ), LDB, WORK( M+1 ), Y,
$ MAX( 1, P ), WORK( M+NP+1 ), LWORK-M-NP, INFO )
WORK( 1 ) = M + NP + MAX( LOPT, INT( WORK( M+NP+1 ) ) )
*
RETURN
*
* End of SGGGLM
*
END
|
