Annotation of rpl/lapack/lapack/zsytf2.f, revision 1.13
1.12 bertrand 1: *> \brief \b ZSYTF2 computes the factorization of a real symmetric indefinite matrix, using the diagonal pivoting method (unblocked algorithm).
1.9 bertrand 2: *
3: * =========== DOCUMENTATION ===========
4: *
5: * Online html documentation available at
6: * http://www.netlib.org/lapack/explore-html/
7: *
8: *> \htmlonly
9: *> Download ZSYTF2 + dependencies
10: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/zsytf2.f">
11: *> [TGZ]</a>
12: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/zsytf2.f">
13: *> [ZIP]</a>
14: *> <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/zsytf2.f">
15: *> [TXT]</a>
16: *> \endhtmlonly
17: *
18: * Definition:
19: * ===========
20: *
21: * SUBROUTINE ZSYTF2( UPLO, N, A, LDA, IPIV, INFO )
22: *
23: * .. Scalar Arguments ..
24: * CHARACTER UPLO
25: * INTEGER INFO, LDA, N
26: * ..
27: * .. Array Arguments ..
28: * INTEGER IPIV( * )
29: * COMPLEX*16 A( LDA, * )
30: * ..
31: *
32: *
33: *> \par Purpose:
34: * =============
35: *>
36: *> \verbatim
37: *>
38: *> ZSYTF2 computes the factorization of a complex symmetric matrix A
39: *> using the Bunch-Kaufman diagonal pivoting method:
40: *>
41: *> A = U*D*U**T or A = L*D*L**T
42: *>
43: *> where U (or L) is a product of permutation and unit upper (lower)
44: *> triangular matrices, U**T is the transpose of U, and D is symmetric and
45: *> block diagonal with 1-by-1 and 2-by-2 diagonal blocks.
46: *>
47: *> This is the unblocked version of the algorithm, calling Level 2 BLAS.
48: *> \endverbatim
49: *
50: * Arguments:
51: * ==========
52: *
53: *> \param[in] UPLO
54: *> \verbatim
55: *> UPLO is CHARACTER*1
56: *> Specifies whether the upper or lower triangular part of the
57: *> symmetric matrix A is stored:
58: *> = 'U': Upper triangular
59: *> = 'L': Lower triangular
60: *> \endverbatim
61: *>
62: *> \param[in] N
63: *> \verbatim
64: *> N is INTEGER
65: *> The order of the matrix A. N >= 0.
66: *> \endverbatim
67: *>
68: *> \param[in,out] A
69: *> \verbatim
70: *> A is COMPLEX*16 array, dimension (LDA,N)
71: *> On entry, the symmetric matrix A. If UPLO = 'U', the leading
72: *> n-by-n upper triangular part of A contains the upper
73: *> triangular part of the matrix A, and the strictly lower
74: *> triangular part of A is not referenced. If UPLO = 'L', the
75: *> leading n-by-n lower triangular part of A contains the lower
76: *> triangular part of the matrix A, and the strictly upper
77: *> triangular part of A is not referenced.
78: *>
79: *> On exit, the block diagonal matrix D and the multipliers used
80: *> to obtain the factor U or L (see below for further details).
81: *> \endverbatim
82: *>
83: *> \param[in] LDA
84: *> \verbatim
85: *> LDA is INTEGER
86: *> The leading dimension of the array A. LDA >= max(1,N).
87: *> \endverbatim
88: *>
89: *> \param[out] IPIV
90: *> \verbatim
91: *> IPIV is INTEGER array, dimension (N)
92: *> Details of the interchanges and the block structure of D.
93: *> If IPIV(k) > 0, then rows and columns k and IPIV(k) were
94: *> interchanged and D(k,k) is a 1-by-1 diagonal block.
95: *> If UPLO = 'U' and IPIV(k) = IPIV(k-1) < 0, then rows and
96: *> columns k-1 and -IPIV(k) were interchanged and D(k-1:k,k-1:k)
97: *> is a 2-by-2 diagonal block. If UPLO = 'L' and IPIV(k) =
98: *> IPIV(k+1) < 0, then rows and columns k+1 and -IPIV(k) were
99: *> interchanged and D(k:k+1,k:k+1) is a 2-by-2 diagonal block.
100: *> \endverbatim
101: *>
102: *> \param[out] INFO
103: *> \verbatim
104: *> INFO is INTEGER
105: *> = 0: successful exit
106: *> < 0: if INFO = -k, the k-th argument had an illegal value
107: *> > 0: if INFO = k, D(k,k) is exactly zero. The factorization
108: *> has been completed, but the block diagonal matrix D is
109: *> exactly singular, and division by zero will occur if it
110: *> is used to solve a system of equations.
111: *> \endverbatim
112: *
113: * Authors:
114: * ========
115: *
116: *> \author Univ. of Tennessee
117: *> \author Univ. of California Berkeley
118: *> \author Univ. of Colorado Denver
119: *> \author NAG Ltd.
120: *
1.12 bertrand 121: *> \date September 2012
1.9 bertrand 122: *
123: *> \ingroup complex16SYcomputational
124: *
125: *> \par Further Details:
126: * =====================
127: *>
128: *> \verbatim
129: *>
130: *> If UPLO = 'U', then A = U*D*U**T, where
131: *> U = P(n)*U(n)* ... *P(k)U(k)* ...,
132: *> i.e., U is a product of terms P(k)*U(k), where k decreases from n to
133: *> 1 in steps of 1 or 2, and D is a block diagonal matrix with 1-by-1
134: *> and 2-by-2 diagonal blocks D(k). P(k) is a permutation matrix as
135: *> defined by IPIV(k), and U(k) is a unit upper triangular matrix, such
136: *> that if the diagonal block D(k) is of order s (s = 1 or 2), then
137: *>
138: *> ( I v 0 ) k-s
139: *> U(k) = ( 0 I 0 ) s
140: *> ( 0 0 I ) n-k
141: *> k-s s n-k
142: *>
143: *> If s = 1, D(k) overwrites A(k,k), and v overwrites A(1:k-1,k).
144: *> If s = 2, the upper triangle of D(k) overwrites A(k-1,k-1), A(k-1,k),
145: *> and A(k,k), and v overwrites A(1:k-2,k-1:k).
146: *>
147: *> If UPLO = 'L', then A = L*D*L**T, where
148: *> L = P(1)*L(1)* ... *P(k)*L(k)* ...,
149: *> i.e., L is a product of terms P(k)*L(k), where k increases from 1 to
150: *> n in steps of 1 or 2, and D is a block diagonal matrix with 1-by-1
151: *> and 2-by-2 diagonal blocks D(k). P(k) is a permutation matrix as
152: *> defined by IPIV(k), and L(k) is a unit lower triangular matrix, such
153: *> that if the diagonal block D(k) is of order s (s = 1 or 2), then
154: *>
155: *> ( I 0 0 ) k-1
156: *> L(k) = ( 0 I 0 ) s
157: *> ( 0 v I ) n-k-s+1
158: *> k-1 s n-k-s+1
159: *>
160: *> If s = 1, D(k) overwrites A(k,k), and v overwrites A(k+1:n,k).
161: *> If s = 2, the lower triangle of D(k) overwrites A(k,k), A(k+1,k),
162: *> and A(k+1,k+1), and v overwrites A(k+2:n,k:k+1).
163: *> \endverbatim
164: *
165: *> \par Contributors:
166: * ==================
167: *>
168: *> \verbatim
169: *>
170: *> 09-29-06 - patch from
171: *> Bobby Cheng, MathWorks
172: *>
173: *> Replace l.209 and l.377
174: *> IF( MAX( ABSAKK, COLMAX ).EQ.ZERO ) THEN
175: *> by
176: *> IF( (MAX( ABSAKK, COLMAX ).EQ.ZERO) .OR. DISNAN(ABSAKK) ) THEN
177: *>
178: *> 1-96 - Based on modifications by J. Lewis, Boeing Computer Services
179: *> Company
180: *> \endverbatim
181: *
182: * =====================================================================
1.1 bertrand 183: SUBROUTINE ZSYTF2( UPLO, N, A, LDA, IPIV, INFO )
184: *
1.12 bertrand 185: * -- LAPACK computational routine (version 3.4.2) --
1.1 bertrand 186: * -- LAPACK is a software package provided by Univ. of Tennessee, --
187: * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
1.12 bertrand 188: * September 2012
1.1 bertrand 189: *
190: * .. Scalar Arguments ..
191: CHARACTER UPLO
192: INTEGER INFO, LDA, N
193: * ..
194: * .. Array Arguments ..
195: INTEGER IPIV( * )
196: COMPLEX*16 A( LDA, * )
197: * ..
198: *
199: * =====================================================================
200: *
201: * .. Parameters ..
202: DOUBLE PRECISION ZERO, ONE
203: PARAMETER ( ZERO = 0.0D+0, ONE = 1.0D+0 )
204: DOUBLE PRECISION EIGHT, SEVTEN
205: PARAMETER ( EIGHT = 8.0D+0, SEVTEN = 17.0D+0 )
206: COMPLEX*16 CONE
207: PARAMETER ( CONE = ( 1.0D+0, 0.0D+0 ) )
208: * ..
209: * .. Local Scalars ..
210: LOGICAL UPPER
211: INTEGER I, IMAX, J, JMAX, K, KK, KP, KSTEP
212: DOUBLE PRECISION ABSAKK, ALPHA, COLMAX, ROWMAX
213: COMPLEX*16 D11, D12, D21, D22, R1, T, WK, WKM1, WKP1, Z
214: * ..
215: * .. External Functions ..
216: LOGICAL DISNAN, LSAME
217: INTEGER IZAMAX
218: EXTERNAL DISNAN, LSAME, IZAMAX
219: * ..
220: * .. External Subroutines ..
221: EXTERNAL XERBLA, ZSCAL, ZSWAP, ZSYR
222: * ..
223: * .. Intrinsic Functions ..
224: INTRINSIC ABS, DBLE, DIMAG, MAX, SQRT
225: * ..
226: * .. Statement Functions ..
227: DOUBLE PRECISION CABS1
228: * ..
229: * .. Statement Function definitions ..
230: CABS1( Z ) = ABS( DBLE( Z ) ) + ABS( DIMAG( Z ) )
231: * ..
232: * .. Executable Statements ..
233: *
234: * Test the input parameters.
235: *
236: INFO = 0
237: UPPER = LSAME( UPLO, 'U' )
238: IF( .NOT.UPPER .AND. .NOT.LSAME( UPLO, 'L' ) ) THEN
239: INFO = -1
240: ELSE IF( N.LT.0 ) THEN
241: INFO = -2
242: ELSE IF( LDA.LT.MAX( 1, N ) ) THEN
243: INFO = -4
244: END IF
245: IF( INFO.NE.0 ) THEN
246: CALL XERBLA( 'ZSYTF2', -INFO )
247: RETURN
248: END IF
249: *
250: * Initialize ALPHA for use in choosing pivot block size.
251: *
252: ALPHA = ( ONE+SQRT( SEVTEN ) ) / EIGHT
253: *
254: IF( UPPER ) THEN
255: *
1.8 bertrand 256: * Factorize A as U*D*U**T using the upper triangle of A
1.1 bertrand 257: *
258: * K is the main loop index, decreasing from N to 1 in steps of
259: * 1 or 2
260: *
261: K = N
262: 10 CONTINUE
263: *
264: * If K < 1, exit from loop
265: *
266: IF( K.LT.1 )
267: $ GO TO 70
268: KSTEP = 1
269: *
270: * Determine rows and columns to be interchanged and whether
271: * a 1-by-1 or 2-by-2 pivot block will be used
272: *
273: ABSAKK = CABS1( A( K, K ) )
274: *
275: * IMAX is the row-index of the largest off-diagonal element in
276: * column K, and COLMAX is its absolute value
277: *
278: IF( K.GT.1 ) THEN
279: IMAX = IZAMAX( K-1, A( 1, K ), 1 )
280: COLMAX = CABS1( A( IMAX, K ) )
281: ELSE
282: COLMAX = ZERO
283: END IF
284: *
285: IF( MAX( ABSAKK, COLMAX ).EQ.ZERO .OR. DISNAN(ABSAKK) ) THEN
286: *
1.8 bertrand 287: * Column K is zero or NaN: set INFO and continue
1.1 bertrand 288: *
289: IF( INFO.EQ.0 )
290: $ INFO = K
291: KP = K
292: ELSE
293: IF( ABSAKK.GE.ALPHA*COLMAX ) THEN
294: *
295: * no interchange, use 1-by-1 pivot block
296: *
297: KP = K
298: ELSE
299: *
300: * JMAX is the column-index of the largest off-diagonal
301: * element in row IMAX, and ROWMAX is its absolute value
302: *
303: JMAX = IMAX + IZAMAX( K-IMAX, A( IMAX, IMAX+1 ), LDA )
304: ROWMAX = CABS1( A( IMAX, JMAX ) )
305: IF( IMAX.GT.1 ) THEN
306: JMAX = IZAMAX( IMAX-1, A( 1, IMAX ), 1 )
307: ROWMAX = MAX( ROWMAX, CABS1( A( JMAX, IMAX ) ) )
308: END IF
309: *
310: IF( ABSAKK.GE.ALPHA*COLMAX*( COLMAX / ROWMAX ) ) THEN
311: *
312: * no interchange, use 1-by-1 pivot block
313: *
314: KP = K
315: ELSE IF( CABS1( A( IMAX, IMAX ) ).GE.ALPHA*ROWMAX ) THEN
316: *
317: * interchange rows and columns K and IMAX, use 1-by-1
318: * pivot block
319: *
320: KP = IMAX
321: ELSE
322: *
323: * interchange rows and columns K-1 and IMAX, use 2-by-2
324: * pivot block
325: *
326: KP = IMAX
327: KSTEP = 2
328: END IF
329: END IF
330: *
331: KK = K - KSTEP + 1
332: IF( KP.NE.KK ) THEN
333: *
334: * Interchange rows and columns KK and KP in the leading
335: * submatrix A(1:k,1:k)
336: *
337: CALL ZSWAP( KP-1, A( 1, KK ), 1, A( 1, KP ), 1 )
338: CALL ZSWAP( KK-KP-1, A( KP+1, KK ), 1, A( KP, KP+1 ),
339: $ LDA )
340: T = A( KK, KK )
341: A( KK, KK ) = A( KP, KP )
342: A( KP, KP ) = T
343: IF( KSTEP.EQ.2 ) THEN
344: T = A( K-1, K )
345: A( K-1, K ) = A( KP, K )
346: A( KP, K ) = T
347: END IF
348: END IF
349: *
350: * Update the leading submatrix
351: *
352: IF( KSTEP.EQ.1 ) THEN
353: *
354: * 1-by-1 pivot block D(k): column k now holds
355: *
356: * W(k) = U(k)*D(k)
357: *
358: * where U(k) is the k-th column of U
359: *
360: * Perform a rank-1 update of A(1:k-1,1:k-1) as
361: *
1.8 bertrand 362: * A := A - U(k)*D(k)*U(k)**T = A - W(k)*1/D(k)*W(k)**T
1.1 bertrand 363: *
364: R1 = CONE / A( K, K )
365: CALL ZSYR( UPLO, K-1, -R1, A( 1, K ), 1, A, LDA )
366: *
367: * Store U(k) in column k
368: *
369: CALL ZSCAL( K-1, R1, A( 1, K ), 1 )
370: ELSE
371: *
372: * 2-by-2 pivot block D(k): columns k and k-1 now hold
373: *
374: * ( W(k-1) W(k) ) = ( U(k-1) U(k) )*D(k)
375: *
376: * where U(k) and U(k-1) are the k-th and (k-1)-th columns
377: * of U
378: *
379: * Perform a rank-2 update of A(1:k-2,1:k-2) as
380: *
1.8 bertrand 381: * A := A - ( U(k-1) U(k) )*D(k)*( U(k-1) U(k) )**T
382: * = A - ( W(k-1) W(k) )*inv(D(k))*( W(k-1) W(k) )**T
1.1 bertrand 383: *
384: IF( K.GT.2 ) THEN
385: *
386: D12 = A( K-1, K )
387: D22 = A( K-1, K-1 ) / D12
388: D11 = A( K, K ) / D12
389: T = CONE / ( D11*D22-CONE )
390: D12 = T / D12
391: *
392: DO 30 J = K - 2, 1, -1
393: WKM1 = D12*( D11*A( J, K-1 )-A( J, K ) )
394: WK = D12*( D22*A( J, K )-A( J, K-1 ) )
395: DO 20 I = J, 1, -1
396: A( I, J ) = A( I, J ) - A( I, K )*WK -
397: $ A( I, K-1 )*WKM1
398: 20 CONTINUE
399: A( J, K ) = WK
400: A( J, K-1 ) = WKM1
401: 30 CONTINUE
402: *
403: END IF
404: *
405: END IF
406: END IF
407: *
408: * Store details of the interchanges in IPIV
409: *
410: IF( KSTEP.EQ.1 ) THEN
411: IPIV( K ) = KP
412: ELSE
413: IPIV( K ) = -KP
414: IPIV( K-1 ) = -KP
415: END IF
416: *
417: * Decrease K and return to the start of the main loop
418: *
419: K = K - KSTEP
420: GO TO 10
421: *
422: ELSE
423: *
1.8 bertrand 424: * Factorize A as L*D*L**T using the lower triangle of A
1.1 bertrand 425: *
426: * K is the main loop index, increasing from 1 to N in steps of
427: * 1 or 2
428: *
429: K = 1
430: 40 CONTINUE
431: *
432: * If K > N, exit from loop
433: *
434: IF( K.GT.N )
435: $ GO TO 70
436: KSTEP = 1
437: *
438: * Determine rows and columns to be interchanged and whether
439: * a 1-by-1 or 2-by-2 pivot block will be used
440: *
441: ABSAKK = CABS1( A( K, K ) )
442: *
443: * IMAX is the row-index of the largest off-diagonal element in
444: * column K, and COLMAX is its absolute value
445: *
446: IF( K.LT.N ) THEN
447: IMAX = K + IZAMAX( N-K, A( K+1, K ), 1 )
448: COLMAX = CABS1( A( IMAX, K ) )
449: ELSE
450: COLMAX = ZERO
451: END IF
452: *
453: IF( MAX( ABSAKK, COLMAX ).EQ.ZERO .OR. DISNAN(ABSAKK) ) THEN
454: *
1.8 bertrand 455: * Column K is zero or NaN: set INFO and continue
1.1 bertrand 456: *
457: IF( INFO.EQ.0 )
458: $ INFO = K
459: KP = K
460: ELSE
461: IF( ABSAKK.GE.ALPHA*COLMAX ) THEN
462: *
463: * no interchange, use 1-by-1 pivot block
464: *
465: KP = K
466: ELSE
467: *
468: * JMAX is the column-index of the largest off-diagonal
469: * element in row IMAX, and ROWMAX is its absolute value
470: *
471: JMAX = K - 1 + IZAMAX( IMAX-K, A( IMAX, K ), LDA )
472: ROWMAX = CABS1( A( IMAX, JMAX ) )
473: IF( IMAX.LT.N ) THEN
474: JMAX = IMAX + IZAMAX( N-IMAX, A( IMAX+1, IMAX ), 1 )
475: ROWMAX = MAX( ROWMAX, CABS1( A( JMAX, IMAX ) ) )
476: END IF
477: *
478: IF( ABSAKK.GE.ALPHA*COLMAX*( COLMAX / ROWMAX ) ) THEN
479: *
480: * no interchange, use 1-by-1 pivot block
481: *
482: KP = K
483: ELSE IF( CABS1( A( IMAX, IMAX ) ).GE.ALPHA*ROWMAX ) THEN
484: *
485: * interchange rows and columns K and IMAX, use 1-by-1
486: * pivot block
487: *
488: KP = IMAX
489: ELSE
490: *
491: * interchange rows and columns K+1 and IMAX, use 2-by-2
492: * pivot block
493: *
494: KP = IMAX
495: KSTEP = 2
496: END IF
497: END IF
498: *
499: KK = K + KSTEP - 1
500: IF( KP.NE.KK ) THEN
501: *
502: * Interchange rows and columns KK and KP in the trailing
503: * submatrix A(k:n,k:n)
504: *
505: IF( KP.LT.N )
506: $ CALL ZSWAP( N-KP, A( KP+1, KK ), 1, A( KP+1, KP ), 1 )
507: CALL ZSWAP( KP-KK-1, A( KK+1, KK ), 1, A( KP, KK+1 ),
508: $ LDA )
509: T = A( KK, KK )
510: A( KK, KK ) = A( KP, KP )
511: A( KP, KP ) = T
512: IF( KSTEP.EQ.2 ) THEN
513: T = A( K+1, K )
514: A( K+1, K ) = A( KP, K )
515: A( KP, K ) = T
516: END IF
517: END IF
518: *
519: * Update the trailing submatrix
520: *
521: IF( KSTEP.EQ.1 ) THEN
522: *
523: * 1-by-1 pivot block D(k): column k now holds
524: *
525: * W(k) = L(k)*D(k)
526: *
527: * where L(k) is the k-th column of L
528: *
529: IF( K.LT.N ) THEN
530: *
531: * Perform a rank-1 update of A(k+1:n,k+1:n) as
532: *
1.8 bertrand 533: * A := A - L(k)*D(k)*L(k)**T = A - W(k)*(1/D(k))*W(k)**T
1.1 bertrand 534: *
535: R1 = CONE / A( K, K )
536: CALL ZSYR( UPLO, N-K, -R1, A( K+1, K ), 1,
537: $ A( K+1, K+1 ), LDA )
538: *
539: * Store L(k) in column K
540: *
541: CALL ZSCAL( N-K, R1, A( K+1, K ), 1 )
542: END IF
543: ELSE
544: *
545: * 2-by-2 pivot block D(k)
546: *
547: IF( K.LT.N-1 ) THEN
548: *
549: * Perform a rank-2 update of A(k+2:n,k+2:n) as
550: *
1.8 bertrand 551: * A := A - ( L(k) L(k+1) )*D(k)*( L(k) L(k+1) )**T
552: * = A - ( W(k) W(k+1) )*inv(D(k))*( W(k) W(k+1) )**T
1.1 bertrand 553: *
554: * where L(k) and L(k+1) are the k-th and (k+1)-th
555: * columns of L
556: *
557: D21 = A( K+1, K )
558: D11 = A( K+1, K+1 ) / D21
559: D22 = A( K, K ) / D21
560: T = CONE / ( D11*D22-CONE )
561: D21 = T / D21
562: *
563: DO 60 J = K + 2, N
564: WK = D21*( D11*A( J, K )-A( J, K+1 ) )
565: WKP1 = D21*( D22*A( J, K+1 )-A( J, K ) )
566: DO 50 I = J, N
567: A( I, J ) = A( I, J ) - A( I, K )*WK -
568: $ A( I, K+1 )*WKP1
569: 50 CONTINUE
570: A( J, K ) = WK
571: A( J, K+1 ) = WKP1
572: 60 CONTINUE
573: END IF
574: END IF
575: END IF
576: *
577: * Store details of the interchanges in IPIV
578: *
579: IF( KSTEP.EQ.1 ) THEN
580: IPIV( K ) = KP
581: ELSE
582: IPIV( K ) = -KP
583: IPIV( K+1 ) = -KP
584: END IF
585: *
586: * Increase K and return to the start of the main loop
587: *
588: K = K + KSTEP
589: GO TO 40
590: *
591: END IF
592: *
593: 70 CONTINUE
594: RETURN
595: *
596: * End of ZSYTF2
597: *
598: END
CVSweb interface <joel.bertrand@systella.fr>