1: *> \brief \b DGEQLF
2: *
3: * =========== DOCUMENTATION ===========
4: *
5: * Online html documentation available at
6: * http://www.netlib.org/lapack/explore-html/
7: *
8: *> \htmlonly
9: *> Download DGEQLF + dependencies
10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dgeqlf.f">
11: *> [TGZ]</a>
12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dgeqlf.f">
13: *> [ZIP]</a>
14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dgeqlf.f">
15: *> [TXT]</a>
16: *> \endhtmlonly
17: *
18: * Definition:
19: * ===========
20: *
21: * SUBROUTINE DGEQLF( M, N, A, LDA, TAU, WORK, LWORK, INFO )
22: *
23: * .. Scalar Arguments ..
24: * INTEGER INFO, LDA, LWORK, M, N
25: * ..
26: * .. Array Arguments ..
27: * DOUBLE PRECISION A( LDA, * ), TAU( * ), WORK( * )
28: * ..
29: *
30: *
31: *> \par Purpose:
32: * =============
33: *>
34: *> \verbatim
35: *>
36: *> DGEQLF computes a QL factorization of a real M-by-N matrix A:
37: *> A = Q * L.
38: *> \endverbatim
39: *
40: * Arguments:
41: * ==========
42: *
43: *> \param[in] M
44: *> \verbatim
45: *> M is INTEGER
46: *> The number of rows of the matrix A. M >= 0.
47: *> \endverbatim
48: *>
49: *> \param[in] N
50: *> \verbatim
51: *> N is INTEGER
52: *> The number of columns of the matrix A. N >= 0.
53: *> \endverbatim
54: *>
55: *> \param[in,out] A
56: *> \verbatim
57: *> A is DOUBLE PRECISION array, dimension (LDA,N)
58: *> On entry, the M-by-N matrix A.
59: *> On exit,
60: *> if m >= n, the lower triangle of the subarray
61: *> A(m-n+1:m,1:n) contains the N-by-N lower triangular matrix L;
62: *> if m <= n, the elements on and below the (n-m)-th
63: *> superdiagonal contain the M-by-N lower trapezoidal matrix L;
64: *> the remaining elements, with the array TAU, represent the
65: *> orthogonal matrix Q as a product of elementary reflectors
66: *> (see Further Details).
67: *> \endverbatim
68: *>
69: *> \param[in] LDA
70: *> \verbatim
71: *> LDA is INTEGER
72: *> The leading dimension of the array A. LDA >= max(1,M).
73: *> \endverbatim
74: *>
75: *> \param[out] TAU
76: *> \verbatim
77: *> TAU is DOUBLE PRECISION array, dimension (min(M,N))
78: *> The scalar factors of the elementary reflectors (see Further
79: *> Details).
80: *> \endverbatim
81: *>
82: *> \param[out] WORK
83: *> \verbatim
84: *> WORK is DOUBLE PRECISION array, dimension (MAX(1,LWORK))
85: *> On exit, if INFO = 0, WORK(1) returns the optimal LWORK.
86: *> \endverbatim
87: *>
88: *> \param[in] LWORK
89: *> \verbatim
90: *> LWORK is INTEGER
91: *> The dimension of the array WORK. LWORK >= max(1,N).
92: *> For optimum performance LWORK >= N*NB, where NB is the
93: *> optimal blocksize.
94: *>
95: *> If LWORK = -1, then a workspace query is assumed; the routine
96: *> only calculates the optimal size of the WORK array, returns
97: *> this value as the first entry of the WORK array, and no error
98: *> message related to LWORK is issued by XERBLA.
99: *> \endverbatim
100: *>
101: *> \param[out] INFO
102: *> \verbatim
103: *> INFO is INTEGER
104: *> = 0: successful exit
105: *> < 0: if INFO = -i, the i-th argument had an illegal value
106: *> \endverbatim
107: *
108: * Authors:
109: * ========
110: *
111: *> \author Univ. of Tennessee
112: *> \author Univ. of California Berkeley
113: *> \author Univ. of Colorado Denver
114: *> \author NAG Ltd.
115: *
116: *> \date November 2011
117: *
118: *> \ingroup doubleGEcomputational
119: *
120: *> \par Further Details:
121: * =====================
122: *>
123: *> \verbatim
124: *>
125: *> The matrix Q is represented as a product of elementary reflectors
126: *>
127: *> Q = H(k) . . . H(2) H(1), where k = min(m,n).
128: *>
129: *> Each H(i) has the form
130: *>
131: *> H(i) = I - tau * v * v**T
132: *>
133: *> where tau is a real scalar, and v is a real vector with
134: *> v(m-k+i+1:m) = 0 and v(m-k+i) = 1; v(1:m-k+i-1) is stored on exit in
135: *> A(1:m-k+i-1,n-k+i), and tau in TAU(i).
136: *> \endverbatim
137: *>
138: * =====================================================================
139: SUBROUTINE DGEQLF( M, N, A, LDA, TAU, WORK, LWORK, INFO )
140: *
141: * -- LAPACK computational routine (version 3.4.0) --
142: * -- LAPACK is a software package provided by Univ. of Tennessee, --
143: * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
144: * November 2011
145: *
146: * .. Scalar Arguments ..
147: INTEGER INFO, LDA, LWORK, M, N
148: * ..
149: * .. Array Arguments ..
150: DOUBLE PRECISION A( LDA, * ), TAU( * ), WORK( * )
151: * ..
152: *
153: * =====================================================================
154: *
155: * .. Local Scalars ..
156: LOGICAL LQUERY
157: INTEGER I, IB, IINFO, IWS, K, KI, KK, LDWORK, LWKOPT,
158: $ MU, NB, NBMIN, NU, NX
159: * ..
160: * .. External Subroutines ..
161: EXTERNAL DGEQL2, DLARFB, DLARFT, XERBLA
162: * ..
163: * .. Intrinsic Functions ..
164: INTRINSIC MAX, MIN
165: * ..
166: * .. External Functions ..
167: INTEGER ILAENV
168: EXTERNAL ILAENV
169: * ..
170: * .. Executable Statements ..
171: *
172: * Test the input arguments
173: *
174: INFO = 0
175: LQUERY = ( LWORK.EQ.-1 )
176: IF( M.LT.0 ) THEN
177: INFO = -1
178: ELSE IF( N.LT.0 ) THEN
179: INFO = -2
180: ELSE IF( LDA.LT.MAX( 1, M ) ) THEN
181: INFO = -4
182: END IF
183: *
184: IF( INFO.EQ.0 ) THEN
185: K = MIN( M, N )
186: IF( K.EQ.0 ) THEN
187: LWKOPT = 1
188: ELSE
189: NB = ILAENV( 1, 'DGEQLF', ' ', M, N, -1, -1 )
190: LWKOPT = N*NB
191: END IF
192: WORK( 1 ) = LWKOPT
193: *
194: IF( LWORK.LT.MAX( 1, N ) .AND. .NOT.LQUERY ) THEN
195: INFO = -7
196: END IF
197: END IF
198: *
199: IF( INFO.NE.0 ) THEN
200: CALL XERBLA( 'DGEQLF', -INFO )
201: RETURN
202: ELSE IF( LQUERY ) THEN
203: RETURN
204: END IF
205: *
206: * Quick return if possible
207: *
208: IF( K.EQ.0 ) THEN
209: RETURN
210: END IF
211: *
212: NBMIN = 2
213: NX = 1
214: IWS = N
215: IF( NB.GT.1 .AND. NB.LT.K ) THEN
216: *
217: * Determine when to cross over from blocked to unblocked code.
218: *
219: NX = MAX( 0, ILAENV( 3, 'DGEQLF', ' ', M, N, -1, -1 ) )
220: IF( NX.LT.K ) THEN
221: *
222: * Determine if workspace is large enough for blocked code.
223: *
224: LDWORK = N
225: IWS = LDWORK*NB
226: IF( LWORK.LT.IWS ) THEN
227: *
228: * Not enough workspace to use optimal NB: reduce NB and
229: * determine the minimum value of NB.
230: *
231: NB = LWORK / LDWORK
232: NBMIN = MAX( 2, ILAENV( 2, 'DGEQLF', ' ', M, N, -1,
233: $ -1 ) )
234: END IF
235: END IF
236: END IF
237: *
238: IF( NB.GE.NBMIN .AND. NB.LT.K .AND. NX.LT.K ) THEN
239: *
240: * Use blocked code initially.
241: * The last kk columns are handled by the block method.
242: *
243: KI = ( ( K-NX-1 ) / NB )*NB
244: KK = MIN( K, KI+NB )
245: *
246: DO 10 I = K - KK + KI + 1, K - KK + 1, -NB
247: IB = MIN( K-I+1, NB )
248: *
249: * Compute the QL factorization of the current block
250: * A(1:m-k+i+ib-1,n-k+i:n-k+i+ib-1)
251: *
252: CALL DGEQL2( M-K+I+IB-1, IB, A( 1, N-K+I ), LDA, TAU( I ),
253: $ WORK, IINFO )
254: IF( N-K+I.GT.1 ) THEN
255: *
256: * Form the triangular factor of the block reflector
257: * H = H(i+ib-1) . . . H(i+1) H(i)
258: *
259: CALL DLARFT( 'Backward', 'Columnwise', M-K+I+IB-1, IB,
260: $ A( 1, N-K+I ), LDA, TAU( I ), WORK, LDWORK )
261: *
262: * Apply H**T to A(1:m-k+i+ib-1,1:n-k+i-1) from the left
263: *
264: CALL DLARFB( 'Left', 'Transpose', 'Backward',
265: $ 'Columnwise', M-K+I+IB-1, N-K+I-1, IB,
266: $ A( 1, N-K+I ), LDA, WORK, LDWORK, A, LDA,
267: $ WORK( IB+1 ), LDWORK )
268: END IF
269: 10 CONTINUE
270: MU = M - K + I + NB - 1
271: NU = N - K + I + NB - 1
272: ELSE
273: MU = M
274: NU = N
275: END IF
276: *
277: * Use unblocked code to factor the last or only block
278: *
279: IF( MU.GT.0 .AND. NU.GT.0 )
280: $ CALL DGEQL2( MU, NU, A, LDA, TAU, WORK, IINFO )
281: *
282: WORK( 1 ) = IWS
283: RETURN
284: *
285: * End of DGEQLF
286: *
287: END
CVSweb interface <joel.bertrand@systella.fr>