1: *> \brief \b ZTZRZF
2: *
3: * =========== DOCUMENTATION ===========
4: *
5: * Online html documentation available at
6: * http://www.netlib.org/lapack/explore-html/
7: *
8: *> \htmlonly
9: *> Download ZTZRZF + dependencies
10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/ztzrzf.f">
11: *> [TGZ]</a>
12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/ztzrzf.f">
13: *> [ZIP]</a>
14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/ztzrzf.f">
15: *> [TXT]</a>
16: *> \endhtmlonly
17: *
18: * Definition:
19: * ===========
20: *
21: * SUBROUTINE ZTZRZF( M, N, A, LDA, TAU, WORK, LWORK, INFO )
22: *
23: * .. Scalar Arguments ..
24: * INTEGER INFO, LDA, LWORK, M, N
25: * ..
26: * .. Array Arguments ..
27: * COMPLEX*16 A( LDA, * ), TAU( * ), WORK( * )
28: * ..
29: *
30: *
31: *> \par Purpose:
32: * =============
33: *>
34: *> \verbatim
35: *>
36: *> ZTZRZF reduces the M-by-N ( M<=N ) complex upper trapezoidal matrix A
37: *> to upper triangular form by means of unitary transformations.
38: *>
39: *> The upper trapezoidal matrix A is factored as
40: *>
41: *> A = ( R 0 ) * Z,
42: *>
43: *> where Z is an N-by-N unitary matrix and R is an M-by-M upper
44: *> triangular matrix.
45: *> \endverbatim
46: *
47: * Arguments:
48: * ==========
49: *
50: *> \param[in] M
51: *> \verbatim
52: *> M is INTEGER
53: *> The number of rows of the matrix A. M >= 0.
54: *> \endverbatim
55: *>
56: *> \param[in] N
57: *> \verbatim
58: *> N is INTEGER
59: *> The number of columns of the matrix A. N >= M.
60: *> \endverbatim
61: *>
62: *> \param[in,out] A
63: *> \verbatim
64: *> A is COMPLEX*16 array, dimension (LDA,N)
65: *> On entry, the leading M-by-N upper trapezoidal part of the
66: *> array A must contain the matrix to be factorized.
67: *> On exit, the leading M-by-M upper triangular part of A
68: *> contains the upper triangular matrix R, and elements M+1 to
69: *> N of the first M rows of A, with the array TAU, represent the
70: *> unitary matrix Z as a product of M elementary reflectors.
71: *> \endverbatim
72: *>
73: *> \param[in] LDA
74: *> \verbatim
75: *> LDA is INTEGER
76: *> The leading dimension of the array A. LDA >= max(1,M).
77: *> \endverbatim
78: *>
79: *> \param[out] TAU
80: *> \verbatim
81: *> TAU is COMPLEX*16 array, dimension (M)
82: *> The scalar factors of the elementary reflectors.
83: *> \endverbatim
84: *>
85: *> \param[out] WORK
86: *> \verbatim
87: *> WORK is COMPLEX*16 array, dimension (MAX(1,LWORK))
88: *> On exit, if INFO = 0, WORK(1) returns the optimal LWORK.
89: *> \endverbatim
90: *>
91: *> \param[in] LWORK
92: *> \verbatim
93: *> LWORK is INTEGER
94: *> The dimension of the array WORK. LWORK >= max(1,M).
95: *> For optimum performance LWORK >= M*NB, where NB is
96: *> the optimal blocksize.
97: *>
98: *> If LWORK = -1, then a workspace query is assumed; the routine
99: *> only calculates the optimal size of the WORK array, returns
100: *> this value as the first entry of the WORK array, and no error
101: *> message related to LWORK is issued by XERBLA.
102: *> \endverbatim
103: *>
104: *> \param[out] INFO
105: *> \verbatim
106: *> INFO is INTEGER
107: *> = 0: successful exit
108: *> < 0: if INFO = -i, the i-th argument had an illegal value
109: *> \endverbatim
110: *
111: * Authors:
112: * ========
113: *
114: *> \author Univ. of Tennessee
115: *> \author Univ. of California Berkeley
116: *> \author Univ. of Colorado Denver
117: *> \author NAG Ltd.
118: *
119: *> \date November 2011
120: *
121: *> \ingroup complex16OTHERcomputational
122: *
123: *> \par Contributors:
124: * ==================
125: *>
126: *> A. Petitet, Computer Science Dept., Univ. of Tenn., Knoxville, USA
127: *
128: *> \par Further Details:
129: * =====================
130: *>
131: *> \verbatim
132: *>
133: *> The factorization is obtained by Householder's method. The kth
134: *> transformation matrix, Z( k ), which is used to introduce zeros into
135: *> the ( m - k + 1 )th row of A, is given in the form
136: *>
137: *> Z( k ) = ( I 0 ),
138: *> ( 0 T( k ) )
139: *>
140: *> where
141: *>
142: *> T( k ) = I - tau*u( k )*u( k )**H, u( k ) = ( 1 ),
143: *> ( 0 )
144: *> ( z( k ) )
145: *>
146: *> tau is a scalar and z( k ) is an ( n - m ) element vector.
147: *> tau and z( k ) are chosen to annihilate the elements of the kth row
148: *> of X.
149: *>
150: *> The scalar tau is returned in the kth element of TAU and the vector
151: *> u( k ) in the kth row of A, such that the elements of z( k ) are
152: *> in a( k, m + 1 ), ..., a( k, n ). The elements of R are returned in
153: *> the upper triangular part of A.
154: *>
155: *> Z is given by
156: *>
157: *> Z = Z( 1 ) * Z( 2 ) * ... * Z( m ).
158: *> \endverbatim
159: *>
160: * =====================================================================
161: SUBROUTINE ZTZRZF( M, N, A, LDA, TAU, WORK, LWORK, INFO )
162: *
163: * -- LAPACK computational routine (version 3.4.0) --
164: * -- LAPACK is a software package provided by Univ. of Tennessee, --
165: * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
166: * November 2011
167: *
168: * .. Scalar Arguments ..
169: INTEGER INFO, LDA, LWORK, M, N
170: * ..
171: * .. Array Arguments ..
172: COMPLEX*16 A( LDA, * ), TAU( * ), WORK( * )
173: * ..
174: *
175: * =====================================================================
176: *
177: * .. Parameters ..
178: COMPLEX*16 ZERO
179: PARAMETER ( ZERO = ( 0.0D+0, 0.0D+0 ) )
180: * ..
181: * .. Local Scalars ..
182: LOGICAL LQUERY
183: INTEGER I, IB, IWS, KI, KK, LDWORK, LWKMIN, LWKOPT,
184: $ M1, MU, NB, NBMIN, NX
185: * ..
186: * .. External Subroutines ..
187: EXTERNAL XERBLA, ZLARZB, ZLARZT, ZLATRZ
188: * ..
189: * .. Intrinsic Functions ..
190: INTRINSIC MAX, MIN
191: * ..
192: * .. External Functions ..
193: INTEGER ILAENV
194: EXTERNAL ILAENV
195: * ..
196: * .. Executable Statements ..
197: *
198: * Test the input arguments
199: *
200: INFO = 0
201: LQUERY = ( LWORK.EQ.-1 )
202: IF( M.LT.0 ) THEN
203: INFO = -1
204: ELSE IF( N.LT.M ) THEN
205: INFO = -2
206: ELSE IF( LDA.LT.MAX( 1, M ) ) THEN
207: INFO = -4
208: END IF
209: *
210: IF( INFO.EQ.0 ) THEN
211: IF( M.EQ.0 .OR. M.EQ.N ) THEN
212: LWKOPT = 1
213: LWKMIN = 1
214: ELSE
215: *
216: * Determine the block size.
217: *
218: NB = ILAENV( 1, 'ZGERQF', ' ', M, N, -1, -1 )
219: LWKOPT = M*NB
220: LWKMIN = MAX( 1, M )
221: END IF
222: WORK( 1 ) = LWKOPT
223: *
224: IF( LWORK.LT.LWKMIN .AND. .NOT.LQUERY ) THEN
225: INFO = -7
226: END IF
227: END IF
228: *
229: IF( INFO.NE.0 ) THEN
230: CALL XERBLA( 'ZTZRZF', -INFO )
231: RETURN
232: ELSE IF( LQUERY ) THEN
233: RETURN
234: END IF
235: *
236: * Quick return if possible
237: *
238: IF( M.EQ.0 ) THEN
239: RETURN
240: ELSE IF( M.EQ.N ) THEN
241: DO 10 I = 1, N
242: TAU( I ) = ZERO
243: 10 CONTINUE
244: RETURN
245: END IF
246: *
247: NBMIN = 2
248: NX = 1
249: IWS = M
250: IF( NB.GT.1 .AND. NB.LT.M ) THEN
251: *
252: * Determine when to cross over from blocked to unblocked code.
253: *
254: NX = MAX( 0, ILAENV( 3, 'ZGERQF', ' ', M, N, -1, -1 ) )
255: IF( NX.LT.M ) THEN
256: *
257: * Determine if workspace is large enough for blocked code.
258: *
259: LDWORK = M
260: IWS = LDWORK*NB
261: IF( LWORK.LT.IWS ) THEN
262: *
263: * Not enough workspace to use optimal NB: reduce NB and
264: * determine the minimum value of NB.
265: *
266: NB = LWORK / LDWORK
267: NBMIN = MAX( 2, ILAENV( 2, 'ZGERQF', ' ', M, N, -1,
268: $ -1 ) )
269: END IF
270: END IF
271: END IF
272: *
273: IF( NB.GE.NBMIN .AND. NB.LT.M .AND. NX.LT.M ) THEN
274: *
275: * Use blocked code initially.
276: * The last kk rows are handled by the block method.
277: *
278: M1 = MIN( M+1, N )
279: KI = ( ( M-NX-1 ) / NB )*NB
280: KK = MIN( M, KI+NB )
281: *
282: DO 20 I = M - KK + KI + 1, M - KK + 1, -NB
283: IB = MIN( M-I+1, NB )
284: *
285: * Compute the TZ factorization of the current block
286: * A(i:i+ib-1,i:n)
287: *
288: CALL ZLATRZ( IB, N-I+1, N-M, A( I, I ), LDA, TAU( I ),
289: $ WORK )
290: IF( I.GT.1 ) THEN
291: *
292: * Form the triangular factor of the block reflector
293: * H = H(i+ib-1) . . . H(i+1) H(i)
294: *
295: CALL ZLARZT( 'Backward', 'Rowwise', N-M, IB, A( I, M1 ),
296: $ LDA, TAU( I ), WORK, LDWORK )
297: *
298: * Apply H to A(1:i-1,i:n) from the right
299: *
300: CALL ZLARZB( 'Right', 'No transpose', 'Backward',
301: $ 'Rowwise', I-1, N-I+1, IB, N-M, A( I, M1 ),
302: $ LDA, WORK, LDWORK, A( 1, I ), LDA,
303: $ WORK( IB+1 ), LDWORK )
304: END IF
305: 20 CONTINUE
306: MU = I + NB - 1
307: ELSE
308: MU = M
309: END IF
310: *
311: * Use unblocked code to factor the last or only block
312: *
313: IF( MU.GT.0 )
314: $ CALL ZLATRZ( MU, N, N-M, A, LDA, TAU, WORK )
315: *
316: WORK( 1 ) = LWKOPT
317: *
318: RETURN
319: *
320: * End of ZTZRZF
321: *
322: END
CVSweb interface <joel.bertrand@systella.fr>