diff options
author | shmel1k <shmel1k@ydb.tech> | 2022-09-02 12:44:59 +0300 |
---|---|---|
committer | shmel1k <shmel1k@ydb.tech> | 2022-09-02 12:44:59 +0300 |
commit | 90d450f74722da7859d6f510a869f6c6908fd12f (patch) | |
tree | 538c718dedc76cdfe37ad6d01ff250dd930d9278 /contrib/libs/clapack/zgerfs.c | |
parent | 01f64c1ecd0d4ffa9e3a74478335f1745f26cc75 (diff) | |
download | ydb-90d450f74722da7859d6f510a869f6c6908fd12f.tar.gz |
[] add metering mode to CLI
Diffstat (limited to 'contrib/libs/clapack/zgerfs.c')
-rw-r--r-- | contrib/libs/clapack/zgerfs.c | 461 |
1 files changed, 461 insertions, 0 deletions
diff --git a/contrib/libs/clapack/zgerfs.c b/contrib/libs/clapack/zgerfs.c new file mode 100644 index 0000000000..4fb472e0b1 --- /dev/null +++ b/contrib/libs/clapack/zgerfs.c @@ -0,0 +1,461 @@ +/* zgerfs.f -- translated by f2c (version 20061008). + You must link the resulting object file with libf2c: + on Microsoft Windows system, link with libf2c.lib; + on Linux or Unix systems, link with .../path/to/libf2c.a -lm + or, if you install libf2c.a in a standard place, with -lf2c -lm + -- in that order, at the end of the command line, as in + cc *.o -lf2c -lm + Source for libf2c is in /netlib/f2c/libf2c.zip, e.g., + + http://www.netlib.org/f2c/libf2c.zip +*/ + +#include "f2c.h" +#include "blaswrap.h" + +/* Table of constant values */ + +static doublecomplex c_b1 = {1.,0.}; +static integer c__1 = 1; + +/* Subroutine */ int zgerfs_(char *trans, integer *n, integer *nrhs, + doublecomplex *a, integer *lda, doublecomplex *af, integer *ldaf, + integer *ipiv, doublecomplex *b, integer *ldb, doublecomplex *x, + integer *ldx, doublereal *ferr, doublereal *berr, doublecomplex *work, + doublereal *rwork, integer *info) +{ + /* System generated locals */ + integer a_dim1, a_offset, af_dim1, af_offset, b_dim1, b_offset, x_dim1, + x_offset, i__1, i__2, i__3, i__4, i__5; + doublereal d__1, d__2, d__3, d__4; + doublecomplex z__1; + + /* Builtin functions */ + double d_imag(doublecomplex *); + + /* Local variables */ + integer i__, j, k; + doublereal s, xk; + integer nz; + doublereal eps; + integer kase; + doublereal safe1, safe2; + extern logical lsame_(char *, char *); + integer isave[3], count; + extern /* Subroutine */ int zgemv_(char *, integer *, integer *, + doublecomplex *, doublecomplex *, integer *, doublecomplex *, + integer *, doublecomplex *, doublecomplex *, integer *), + zcopy_(integer *, doublecomplex *, integer *, doublecomplex *, + integer *), zaxpy_(integer *, doublecomplex *, doublecomplex *, + integer *, doublecomplex *, integer *), zlacn2_(integer *, + doublecomplex *, doublecomplex *, doublereal *, integer *, + integer *); + extern doublereal dlamch_(char *); + doublereal safmin; + extern /* Subroutine */ int xerbla_(char *, integer *); + logical notran; + char transn[1], transt[1]; + doublereal lstres; + extern /* Subroutine */ int zgetrs_(char *, integer *, integer *, + doublecomplex *, integer *, integer *, doublecomplex *, integer *, + integer *); + + +/* -- LAPACK routine (version 3.2) -- */ +/* Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. */ +/* November 2006 */ + +/* Modified to call ZLACN2 in place of ZLACON, 10 Feb 03, SJH. */ + +/* .. Scalar Arguments .. */ +/* .. */ +/* .. Array Arguments .. */ +/* .. */ + +/* Purpose */ +/* ======= */ + +/* ZGERFS improves the computed solution to a system of linear */ +/* equations and provides error bounds and backward error estimates for */ +/* the solution. */ + +/* Arguments */ +/* ========= */ + +/* TRANS (input) CHARACTER*1 */ +/* Specifies the form of the system of equations: */ +/* = 'N': A * X = B (No transpose) */ +/* = 'T': A**T * X = B (Transpose) */ +/* = 'C': A**H * X = B (Conjugate transpose) */ + +/* N (input) INTEGER */ +/* The order of the matrix A. N >= 0. */ + +/* NRHS (input) INTEGER */ +/* The number of right hand sides, i.e., the number of columns */ +/* of the matrices B and X. NRHS >= 0. */ + +/* A (input) COMPLEX*16 array, dimension (LDA,N) */ +/* The original N-by-N matrix A. */ + +/* LDA (input) INTEGER */ +/* The leading dimension of the array A. LDA >= max(1,N). */ + +/* AF (input) COMPLEX*16 array, dimension (LDAF,N) */ +/* The factors L and U from the factorization A = P*L*U */ +/* as computed by ZGETRF. */ + +/* LDAF (input) INTEGER */ +/* The leading dimension of the array AF. LDAF >= max(1,N). */ + +/* IPIV (input) INTEGER array, dimension (N) */ +/* The pivot indices from ZGETRF; for 1<=i<=N, row i of the */ +/* matrix was interchanged with row IPIV(i). */ + +/* B (input) COMPLEX*16 array, dimension (LDB,NRHS) */ +/* The right hand side matrix B. */ + +/* LDB (input) INTEGER */ +/* The leading dimension of the array B. LDB >= max(1,N). */ + +/* X (input/output) COMPLEX*16 array, dimension (LDX,NRHS) */ +/* On entry, the solution matrix X, as computed by ZGETRS. */ +/* On exit, the improved solution matrix X. */ + +/* LDX (input) INTEGER */ +/* The leading dimension of the array X. LDX >= max(1,N). */ + +/* FERR (output) DOUBLE PRECISION array, dimension (NRHS) */ +/* The estimated forward error bound for each solution vector */ +/* X(j) (the j-th column of the solution matrix X). */ +/* If XTRUE is the true solution corresponding to X(j), FERR(j) */ +/* is an estimated upper bound for the magnitude of the largest */ +/* element in (X(j) - XTRUE) divided by the magnitude of the */ +/* largest element in X(j). The estimate is as reliable as */ +/* the estimate for RCOND, and is almost always a slight */ +/* overestimate of the true error. */ + +/* BERR (output) DOUBLE PRECISION array, dimension (NRHS) */ +/* The componentwise relative backward error of each solution */ +/* vector X(j) (i.e., the smallest relative change in */ +/* any element of A or B that makes X(j) an exact solution). */ + +/* WORK (workspace) COMPLEX*16 array, dimension (2*N) */ + +/* RWORK (workspace) DOUBLE PRECISION array, dimension (N) */ + +/* INFO (output) INTEGER */ +/* = 0: successful exit */ +/* < 0: if INFO = -i, the i-th argument had an illegal value */ + +/* Internal Parameters */ +/* =================== */ + +/* ITMAX is the maximum number of steps of iterative refinement. */ + +/* ===================================================================== */ + +/* .. Parameters .. */ +/* .. */ +/* .. Local Scalars .. */ +/* .. */ +/* .. Local Arrays .. */ +/* .. */ +/* .. External Functions .. */ +/* .. */ +/* .. External Subroutines .. */ +/* .. */ +/* .. Intrinsic Functions .. */ +/* .. */ +/* .. Statement Functions .. */ +/* .. */ +/* .. Statement Function definitions .. */ +/* .. */ +/* .. Executable Statements .. */ + +/* Test the input parameters. */ + + /* Parameter adjustments */ + a_dim1 = *lda; + a_offset = 1 + a_dim1; + a -= a_offset; + af_dim1 = *ldaf; + af_offset = 1 + af_dim1; + af -= af_offset; + --ipiv; + b_dim1 = *ldb; + b_offset = 1 + b_dim1; + b -= b_offset; + x_dim1 = *ldx; + x_offset = 1 + x_dim1; + x -= x_offset; + --ferr; + --berr; + --work; + --rwork; + + /* Function Body */ + *info = 0; + notran = lsame_(trans, "N"); + if (! notran && ! lsame_(trans, "T") && ! lsame_( + trans, "C")) { + *info = -1; + } else if (*n < 0) { + *info = -2; + } else if (*nrhs < 0) { + *info = -3; + } else if (*lda < max(1,*n)) { + *info = -5; + } else if (*ldaf < max(1,*n)) { + *info = -7; + } else if (*ldb < max(1,*n)) { + *info = -10; + } else if (*ldx < max(1,*n)) { + *info = -12; + } + if (*info != 0) { + i__1 = -(*info); + xerbla_("ZGERFS", &i__1); + return 0; + } + +/* Quick return if possible */ + + if (*n == 0 || *nrhs == 0) { + i__1 = *nrhs; + for (j = 1; j <= i__1; ++j) { + ferr[j] = 0.; + berr[j] = 0.; +/* L10: */ + } + return 0; + } + + if (notran) { + *(unsigned char *)transn = 'N'; + *(unsigned char *)transt = 'C'; + } else { + *(unsigned char *)transn = 'C'; + *(unsigned char *)transt = 'N'; + } + +/* NZ = maximum number of nonzero elements in each row of A, plus 1 */ + + nz = *n + 1; + eps = dlamch_("Epsilon"); + safmin = dlamch_("Safe minimum"); + safe1 = nz * safmin; + safe2 = safe1 / eps; + +/* Do for each right hand side */ + + i__1 = *nrhs; + for (j = 1; j <= i__1; ++j) { + + count = 1; + lstres = 3.; +L20: + +/* Loop until stopping criterion is satisfied. */ + +/* Compute residual R = B - op(A) * X, */ +/* where op(A) = A, A**T, or A**H, depending on TRANS. */ + + zcopy_(n, &b[j * b_dim1 + 1], &c__1, &work[1], &c__1); + z__1.r = -1., z__1.i = -0.; + zgemv_(trans, n, n, &z__1, &a[a_offset], lda, &x[j * x_dim1 + 1], & + c__1, &c_b1, &work[1], &c__1); + +/* Compute componentwise relative backward error from formula */ + +/* max(i) ( abs(R(i)) / ( abs(op(A))*abs(X) + abs(B) )(i) ) */ + +/* where abs(Z) is the componentwise absolute value of the matrix */ +/* or vector Z. If the i-th component of the denominator is less */ +/* than SAFE2, then SAFE1 is added to the i-th components of the */ +/* numerator and denominator before dividing. */ + + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { + i__3 = i__ + j * b_dim1; + rwork[i__] = (d__1 = b[i__3].r, abs(d__1)) + (d__2 = d_imag(&b[ + i__ + j * b_dim1]), abs(d__2)); +/* L30: */ + } + +/* Compute abs(op(A))*abs(X) + abs(B). */ + + if (notran) { + i__2 = *n; + for (k = 1; k <= i__2; ++k) { + i__3 = k + j * x_dim1; + xk = (d__1 = x[i__3].r, abs(d__1)) + (d__2 = d_imag(&x[k + j * + x_dim1]), abs(d__2)); + i__3 = *n; + for (i__ = 1; i__ <= i__3; ++i__) { + i__4 = i__ + k * a_dim1; + rwork[i__] += ((d__1 = a[i__4].r, abs(d__1)) + (d__2 = + d_imag(&a[i__ + k * a_dim1]), abs(d__2))) * xk; +/* L40: */ + } +/* L50: */ + } + } else { + i__2 = *n; + for (k = 1; k <= i__2; ++k) { + s = 0.; + i__3 = *n; + for (i__ = 1; i__ <= i__3; ++i__) { + i__4 = i__ + k * a_dim1; + i__5 = i__ + j * x_dim1; + s += ((d__1 = a[i__4].r, abs(d__1)) + (d__2 = d_imag(&a[ + i__ + k * a_dim1]), abs(d__2))) * ((d__3 = x[i__5] + .r, abs(d__3)) + (d__4 = d_imag(&x[i__ + j * + x_dim1]), abs(d__4))); +/* L60: */ + } + rwork[k] += s; +/* L70: */ + } + } + s = 0.; + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { + if (rwork[i__] > safe2) { +/* Computing MAX */ + i__3 = i__; + d__3 = s, d__4 = ((d__1 = work[i__3].r, abs(d__1)) + (d__2 = + d_imag(&work[i__]), abs(d__2))) / rwork[i__]; + s = max(d__3,d__4); + } else { +/* Computing MAX */ + i__3 = i__; + d__3 = s, d__4 = ((d__1 = work[i__3].r, abs(d__1)) + (d__2 = + d_imag(&work[i__]), abs(d__2)) + safe1) / (rwork[i__] + + safe1); + s = max(d__3,d__4); + } +/* L80: */ + } + berr[j] = s; + +/* Test stopping criterion. Continue iterating if */ +/* 1) The residual BERR(J) is larger than machine epsilon, and */ +/* 2) BERR(J) decreased by at least a factor of 2 during the */ +/* last iteration, and */ +/* 3) At most ITMAX iterations tried. */ + + if (berr[j] > eps && berr[j] * 2. <= lstres && count <= 5) { + +/* Update solution and try again. */ + + zgetrs_(trans, n, &c__1, &af[af_offset], ldaf, &ipiv[1], &work[1], + n, info); + zaxpy_(n, &c_b1, &work[1], &c__1, &x[j * x_dim1 + 1], &c__1); + lstres = berr[j]; + ++count; + goto L20; + } + +/* Bound error from formula */ + +/* norm(X - XTRUE) / norm(X) .le. FERR = */ +/* norm( abs(inv(op(A)))* */ +/* ( abs(R) + NZ*EPS*( abs(op(A))*abs(X)+abs(B) ))) / norm(X) */ + +/* where */ +/* norm(Z) is the magnitude of the largest component of Z */ +/* inv(op(A)) is the inverse of op(A) */ +/* abs(Z) is the componentwise absolute value of the matrix or */ +/* vector Z */ +/* NZ is the maximum number of nonzeros in any row of A, plus 1 */ +/* EPS is machine epsilon */ + +/* The i-th component of abs(R)+NZ*EPS*(abs(op(A))*abs(X)+abs(B)) */ +/* is incremented by SAFE1 if the i-th component of */ +/* abs(op(A))*abs(X) + abs(B) is less than SAFE2. */ + +/* Use ZLACN2 to estimate the infinity-norm of the matrix */ +/* inv(op(A)) * diag(W), */ +/* where W = abs(R) + NZ*EPS*( abs(op(A))*abs(X)+abs(B) ))) */ + + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { + if (rwork[i__] > safe2) { + i__3 = i__; + rwork[i__] = (d__1 = work[i__3].r, abs(d__1)) + (d__2 = + d_imag(&work[i__]), abs(d__2)) + nz * eps * rwork[i__] + ; + } else { + i__3 = i__; + rwork[i__] = (d__1 = work[i__3].r, abs(d__1)) + (d__2 = + d_imag(&work[i__]), abs(d__2)) + nz * eps * rwork[i__] + + safe1; + } +/* L90: */ + } + + kase = 0; +L100: + zlacn2_(n, &work[*n + 1], &work[1], &ferr[j], &kase, isave); + if (kase != 0) { + if (kase == 1) { + +/* Multiply by diag(W)*inv(op(A)**H). */ + + zgetrs_(transt, n, &c__1, &af[af_offset], ldaf, &ipiv[1], & + work[1], n, info); + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { + i__3 = i__; + i__4 = i__; + i__5 = i__; + z__1.r = rwork[i__4] * work[i__5].r, z__1.i = rwork[i__4] + * work[i__5].i; + work[i__3].r = z__1.r, work[i__3].i = z__1.i; +/* L110: */ + } + } else { + +/* Multiply by inv(op(A))*diag(W). */ + + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { + i__3 = i__; + i__4 = i__; + i__5 = i__; + z__1.r = rwork[i__4] * work[i__5].r, z__1.i = rwork[i__4] + * work[i__5].i; + work[i__3].r = z__1.r, work[i__3].i = z__1.i; +/* L120: */ + } + zgetrs_(transn, n, &c__1, &af[af_offset], ldaf, &ipiv[1], & + work[1], n, info); + } + goto L100; + } + +/* Normalize error. */ + + lstres = 0.; + i__2 = *n; + for (i__ = 1; i__ <= i__2; ++i__) { +/* Computing MAX */ + i__3 = i__ + j * x_dim1; + d__3 = lstres, d__4 = (d__1 = x[i__3].r, abs(d__1)) + (d__2 = + d_imag(&x[i__ + j * x_dim1]), abs(d__2)); + lstres = max(d__3,d__4); +/* L130: */ + } + if (lstres != 0.) { + ferr[j] /= lstres; + } + +/* L140: */ + } + + return 0; + +/* End of ZGERFS */ + +} /* zgerfs_ */ |