![[BACK]](/cvsweb/icons/back.gif){kind=icon}
Return to dtpsv.f CVS log ![[TXT]](/cvsweb/icons/text.gif){kind=icon}
![[DIR]](/cvsweb/icons/dir.gif){kind=icon}
Up to [local] / rpl / lapack / blas|
Return to dtpsv.f CVS log | Up to [local] / rpl / lapack / blas |
1.8 bertrand 1: *> \brief \b DTPSV
2: *
3: * =========== DOCUMENTATION ===========
4: *
1.13 bertrand 5: * Online html documentation available at
6: * http://www.netlib.org/lapack/explore-html/
1.8 bertrand 7: *
8: * Definition:
9: * ===========
10: *
11: * SUBROUTINE DTPSV(UPLO,TRANS,DIAG,N,AP,X,INCX)
1.13 bertrand 12: *
1.8 bertrand 13: * .. Scalar Arguments ..
14: * INTEGER INCX,N
15: * CHARACTER DIAG,TRANS,UPLO
16: * ..
17: * .. Array Arguments ..
18: * DOUBLE PRECISION AP(*),X(*)
19: * ..
1.13 bertrand 20: *
1.8 bertrand 21: *
22: *> \par Purpose:
23: * =============
24: *>
25: *> \verbatim
26: *>
27: *> DTPSV solves one of the systems of equations
28: *>
29: *> A*x = b, or A**T*x = b,
30: *>
31: *> where b and x are n element vectors and A is an n by n unit, or
32: *> non-unit, upper or lower triangular matrix, supplied in packed form.
33: *>
34: *> No test for singularity or near-singularity is included in this
35: *> routine. Such tests must be performed before calling this routine.
36: *> \endverbatim
37: *
38: * Arguments:
39: * ==========
40: *
41: *> \param[in] UPLO
42: *> \verbatim
43: *> UPLO is CHARACTER*1
44: *> On entry, UPLO specifies whether the matrix is an upper or
45: *> lower triangular matrix as follows:
46: *>
47: *> UPLO = 'U' or 'u' A is an upper triangular matrix.
48: *>
49: *> UPLO = 'L' or 'l' A is a lower triangular matrix.
50: *> \endverbatim
51: *>
52: *> \param[in] TRANS
53: *> \verbatim
54: *> TRANS is CHARACTER*1
55: *> On entry, TRANS specifies the equations to be solved as
56: *> follows:
57: *>
58: *> TRANS = 'N' or 'n' A*x = b.
59: *>
60: *> TRANS = 'T' or 't' A**T*x = b.
61: *>
62: *> TRANS = 'C' or 'c' A**T*x = b.
63: *> \endverbatim
64: *>
65: *> \param[in] DIAG
66: *> \verbatim
67: *> DIAG is CHARACTER*1
68: *> On entry, DIAG specifies whether or not A is unit
69: *> triangular as follows:
70: *>
71: *> DIAG = 'U' or 'u' A is assumed to be unit triangular.
72: *>
73: *> DIAG = 'N' or 'n' A is not assumed to be unit
74: *> triangular.
75: *> \endverbatim
76: *>
77: *> \param[in] N
78: *> \verbatim
79: *> N is INTEGER
80: *> On entry, N specifies the order of the matrix A.
81: *> N must be at least zero.
82: *> \endverbatim
83: *>
84: *> \param[in] AP
85: *> \verbatim
1.14 bertrand 86: *> AP is DOUBLE PRECISION array, dimension at least
1.8 bertrand 87: *> ( ( n*( n + 1 ) )/2 ).
88: *> Before entry with UPLO = 'U' or 'u', the array AP must
89: *> contain the upper triangular matrix packed sequentially,
90: *> column by column, so that AP( 1 ) contains a( 1, 1 ),
91: *> AP( 2 ) and AP( 3 ) contain a( 1, 2 ) and a( 2, 2 )
92: *> respectively, and so on.
93: *> Before entry with UPLO = 'L' or 'l', the array AP must
94: *> contain the lower triangular matrix packed sequentially,
95: *> column by column, so that AP( 1 ) contains a( 1, 1 ),
96: *> AP( 2 ) and AP( 3 ) contain a( 2, 1 ) and a( 3, 1 )
97: *> respectively, and so on.
98: *> Note that when DIAG = 'U' or 'u', the diagonal elements of
99: *> A are not referenced, but are assumed to be unity.
100: *> \endverbatim
101: *>
102: *> \param[in,out] X
103: *> \verbatim
1.14 bertrand 104: *> X is DOUBLE PRECISION array, dimension at least
1.8 bertrand 105: *> ( 1 + ( n - 1 )*abs( INCX ) ).
106: *> Before entry, the incremented array X must contain the n
107: *> element right-hand side vector b. On exit, X is overwritten
108: *> with the solution vector x.
109: *> \endverbatim
110: *>
111: *> \param[in] INCX
112: *> \verbatim
113: *> INCX is INTEGER
114: *> On entry, INCX specifies the increment for the elements of
115: *> X. INCX must not be zero.
116: *> \endverbatim
117: *
118: * Authors:
119: * ========
120: *
1.13 bertrand 121: *> \author Univ. of Tennessee
122: *> \author Univ. of California Berkeley
123: *> \author Univ. of Colorado Denver
124: *> \author NAG Ltd.
1.8 bertrand 125: *
1.13 bertrand 126: *> \date December 2016
1.8 bertrand 127: *
128: *> \ingroup double_blas_level2
129: *
130: *> \par Further Details:
131: * =====================
132: *>
133: *> \verbatim
134: *>
135: *> Level 2 Blas routine.
136: *>
137: *> -- Written on 22-October-1986.
138: *> Jack Dongarra, Argonne National Lab.
139: *> Jeremy Du Croz, Nag Central Office.
140: *> Sven Hammarling, Nag Central Office.
141: *> Richard Hanson, Sandia National Labs.
142: *> \endverbatim
143: *>
144: * =====================================================================
1.1 bertrand 145: SUBROUTINE DTPSV(UPLO,TRANS,DIAG,N,AP,X,INCX)
1.8 bertrand 146: *
1.13 bertrand 147: * -- Reference BLAS level2 routine (version 3.7.0) --
1.8 bertrand 148: * -- Reference BLAS is a software package provided by Univ. of Tennessee, --
149: * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
1.13 bertrand 150: * December 2016
1.8 bertrand 151: *
1.1 bertrand 152: * .. Scalar Arguments ..
153: INTEGER INCX,N
154: CHARACTER DIAG,TRANS,UPLO
155: * ..
156: * .. Array Arguments ..
157: DOUBLE PRECISION AP(*),X(*)
158: * ..
159: *
160: * =====================================================================
161: *
162: * .. Parameters ..
163: DOUBLE PRECISION ZERO
164: PARAMETER (ZERO=0.0D+0)
165: * ..
166: * .. Local Scalars ..
167: DOUBLE PRECISION TEMP
168: INTEGER I,INFO,IX,J,JX,K,KK,KX
169: LOGICAL NOUNIT
170: * ..
171: * .. External Functions ..
172: LOGICAL LSAME
173: EXTERNAL LSAME
174: * ..
175: * .. External Subroutines ..
176: EXTERNAL XERBLA
177: * ..
178: *
179: * Test the input parameters.
180: *
181: INFO = 0
182: IF (.NOT.LSAME(UPLO,'U') .AND. .NOT.LSAME(UPLO,'L')) THEN
183: INFO = 1
184: ELSE IF (.NOT.LSAME(TRANS,'N') .AND. .NOT.LSAME(TRANS,'T') .AND.
185: + .NOT.LSAME(TRANS,'C')) THEN
186: INFO = 2
187: ELSE IF (.NOT.LSAME(DIAG,'U') .AND. .NOT.LSAME(DIAG,'N')) THEN
188: INFO = 3
189: ELSE IF (N.LT.0) THEN
190: INFO = 4
191: ELSE IF (INCX.EQ.0) THEN
192: INFO = 7
193: END IF
194: IF (INFO.NE.0) THEN
195: CALL XERBLA('DTPSV ',INFO)
196: RETURN
197: END IF
198: *
199: * Quick return if possible.
200: *
201: IF (N.EQ.0) RETURN
202: *
203: NOUNIT = LSAME(DIAG,'N')
204: *
205: * Set up the start point in X if the increment is not unity. This
206: * will be ( N - 1 )*INCX too small for descending loops.
207: *
208: IF (INCX.LE.0) THEN
209: KX = 1 - (N-1)*INCX
210: ELSE IF (INCX.NE.1) THEN
211: KX = 1
212: END IF
213: *
214: * Start the operations. In this version the elements of AP are
215: * accessed sequentially with one pass through AP.
216: *
217: IF (LSAME(TRANS,'N')) THEN
218: *
219: * Form x := inv( A )*x.
220: *
221: IF (LSAME(UPLO,'U')) THEN
222: KK = (N* (N+1))/2
223: IF (INCX.EQ.1) THEN
224: DO 20 J = N,1,-1
225: IF (X(J).NE.ZERO) THEN
226: IF (NOUNIT) X(J) = X(J)/AP(KK)
227: TEMP = X(J)
228: K = KK - 1
229: DO 10 I = J - 1,1,-1
230: X(I) = X(I) - TEMP*AP(K)
231: K = K - 1
232: 10 CONTINUE
233: END IF
234: KK = KK - J
235: 20 CONTINUE
236: ELSE
237: JX = KX + (N-1)*INCX
238: DO 40 J = N,1,-1
239: IF (X(JX).NE.ZERO) THEN
240: IF (NOUNIT) X(JX) = X(JX)/AP(KK)
241: TEMP = X(JX)
242: IX = JX
243: DO 30 K = KK - 1,KK - J + 1,-1
244: IX = IX - INCX
245: X(IX) = X(IX) - TEMP*AP(K)
246: 30 CONTINUE
247: END IF
248: JX = JX - INCX
249: KK = KK - J
250: 40 CONTINUE
251: END IF
252: ELSE
253: KK = 1
254: IF (INCX.EQ.1) THEN
255: DO 60 J = 1,N
256: IF (X(J).NE.ZERO) THEN
257: IF (NOUNIT) X(J) = X(J)/AP(KK)
258: TEMP = X(J)
259: K = KK + 1
260: DO 50 I = J + 1,N
261: X(I) = X(I) - TEMP*AP(K)
262: K = K + 1
263: 50 CONTINUE
264: END IF
265: KK = KK + (N-J+1)
266: 60 CONTINUE
267: ELSE
268: JX = KX
269: DO 80 J = 1,N
270: IF (X(JX).NE.ZERO) THEN
271: IF (NOUNIT) X(JX) = X(JX)/AP(KK)
272: TEMP = X(JX)
273: IX = JX
274: DO 70 K = KK + 1,KK + N - J
275: IX = IX + INCX
276: X(IX) = X(IX) - TEMP*AP(K)
277: 70 CONTINUE
278: END IF
279: JX = JX + INCX
280: KK = KK + (N-J+1)
281: 80 CONTINUE
282: END IF
283: END IF
284: ELSE
285: *
1.7 bertrand 286: * Form x := inv( A**T )*x.
1.1 bertrand 287: *
288: IF (LSAME(UPLO,'U')) THEN
289: KK = 1
290: IF (INCX.EQ.1) THEN
291: DO 100 J = 1,N
292: TEMP = X(J)
293: K = KK
294: DO 90 I = 1,J - 1
295: TEMP = TEMP - AP(K)*X(I)
296: K = K + 1
297: 90 CONTINUE
298: IF (NOUNIT) TEMP = TEMP/AP(KK+J-1)
299: X(J) = TEMP
300: KK = KK + J
301: 100 CONTINUE
302: ELSE
303: JX = KX
304: DO 120 J = 1,N
305: TEMP = X(JX)
306: IX = KX
307: DO 110 K = KK,KK + J - 2
308: TEMP = TEMP - AP(K)*X(IX)
309: IX = IX + INCX
310: 110 CONTINUE
311: IF (NOUNIT) TEMP = TEMP/AP(KK+J-1)
312: X(JX) = TEMP
313: JX = JX + INCX
314: KK = KK + J
315: 120 CONTINUE
316: END IF
317: ELSE
318: KK = (N* (N+1))/2
319: IF (INCX.EQ.1) THEN
320: DO 140 J = N,1,-1
321: TEMP = X(J)
322: K = KK
323: DO 130 I = N,J + 1,-1
324: TEMP = TEMP - AP(K)*X(I)
325: K = K - 1
326: 130 CONTINUE
327: IF (NOUNIT) TEMP = TEMP/AP(KK-N+J)
328: X(J) = TEMP
329: KK = KK - (N-J+1)
330: 140 CONTINUE
331: ELSE
332: KX = KX + (N-1)*INCX
333: JX = KX
334: DO 160 J = N,1,-1
335: TEMP = X(JX)
336: IX = KX
337: DO 150 K = KK,KK - (N- (J+1)),-1
338: TEMP = TEMP - AP(K)*X(IX)
339: IX = IX - INCX
340: 150 CONTINUE
341: IF (NOUNIT) TEMP = TEMP/AP(KK-N+J)
342: X(JX) = TEMP
343: JX = JX - INCX
344: KK = KK - (N-J+1)
345: 160 CONTINUE
346: END IF
347: END IF
348: END IF
349: *
350: RETURN
351: *
352: * End of DTPSV .
353: *
354: END