slaed1.c
Go to the documentation of this file.
00001 /* slaed1.f -- translated by f2c (version 20061008).
00002    You must link the resulting object file with libf2c:
00003         on Microsoft Windows system, link with libf2c.lib;
00004         on Linux or Unix systems, link with .../path/to/libf2c.a -lm
00005         or, if you install libf2c.a in a standard place, with -lf2c -lm
00006         -- in that order, at the end of the command line, as in
00007                 cc *.o -lf2c -lm
00008         Source for libf2c is in /netlib/f2c/libf2c.zip, e.g.,
00009 
00010                 http://www.netlib.org/f2c/libf2c.zip
00011 */
00012 
00013 #include "f2c.h"
00014 #include "blaswrap.h"
00015 
00016 /* Table of constant values */
00017 
00018 static integer c__1 = 1;
00019 static integer c_n1 = -1;
00020 
00021 /* Subroutine */ int slaed1_(integer *n, real *d__, real *q, integer *ldq, 
00022         integer *indxq, real *rho, integer *cutpnt, real *work, integer *
00023         iwork, integer *info)
00024 {
00025     /* System generated locals */
00026     integer q_dim1, q_offset, i__1, i__2;
00027 
00028     /* Local variables */
00029     integer i__, k, n1, n2, is, iw, iz, iq2, cpp1, indx, indxc, indxp;
00030     extern /* Subroutine */ int scopy_(integer *, real *, integer *, real *, 
00031             integer *), slaed2_(integer *, integer *, integer *, real *, real 
00032             *, integer *, integer *, real *, real *, real *, real *, real *, 
00033             integer *, integer *, integer *, integer *, integer *), slaed3_(
00034             integer *, integer *, integer *, real *, real *, integer *, real *
00035 , real *, real *, integer *, integer *, real *, real *, integer *)
00036             ;
00037     integer idlmda;
00038     extern /* Subroutine */ int xerbla_(char *, integer *), slamrg_(
00039             integer *, integer *, real *, integer *, integer *, integer *);
00040     integer coltyp;
00041 
00042 
00043 /*  -- LAPACK routine (version 3.2) -- */
00044 /*     Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. */
00045 /*     November 2006 */
00046 
00047 /*     .. Scalar Arguments .. */
00048 /*     .. */
00049 /*     .. Array Arguments .. */
00050 /*     .. */
00051 
00052 /*  Purpose */
00053 /*  ======= */
00054 
00055 /*  SLAED1 computes the updated eigensystem of a diagonal */
00056 /*  matrix after modification by a rank-one symmetric matrix.  This */
00057 /*  routine is used only for the eigenproblem which requires all */
00058 /*  eigenvalues and eigenvectors of a tridiagonal matrix.  SLAED7 handles */
00059 /*  the case in which eigenvalues only or eigenvalues and eigenvectors */
00060 /*  of a full symmetric matrix (which was reduced to tridiagonal form) */
00061 /*  are desired. */
00062 
00063 /*    T = Q(in) ( D(in) + RHO * Z*Z' ) Q'(in) = Q(out) * D(out) * Q'(out) */
00064 
00065 /*     where Z = Q'u, u is a vector of length N with ones in the */
00066 /*     CUTPNT and CUTPNT + 1 th elements and zeros elsewhere. */
00067 
00068 /*     The eigenvectors of the original matrix are stored in Q, and the */
00069 /*     eigenvalues are in D.  The algorithm consists of three stages: */
00070 
00071 /*        The first stage consists of deflating the size of the problem */
00072 /*        when there are multiple eigenvalues or if there is a zero in */
00073 /*        the Z vector.  For each such occurence the dimension of the */
00074 /*        secular equation problem is reduced by one.  This stage is */
00075 /*        performed by the routine SLAED2. */
00076 
00077 /*        The second stage consists of calculating the updated */
00078 /*        eigenvalues. This is done by finding the roots of the secular */
00079 /*        equation via the routine SLAED4 (as called by SLAED3). */
00080 /*        This routine also calculates the eigenvectors of the current */
00081 /*        problem. */
00082 
00083 /*        The final stage consists of computing the updated eigenvectors */
00084 /*        directly using the updated eigenvalues.  The eigenvectors for */
00085 /*        the current problem are multiplied with the eigenvectors from */
00086 /*        the overall problem. */
00087 
00088 /*  Arguments */
00089 /*  ========= */
00090 
00091 /*  N      (input) INTEGER */
00092 /*         The dimension of the symmetric tridiagonal matrix.  N >= 0. */
00093 
00094 /*  D      (input/output) REAL array, dimension (N) */
00095 /*         On entry, the eigenvalues of the rank-1-perturbed matrix. */
00096 /*         On exit, the eigenvalues of the repaired matrix. */
00097 
00098 /*  Q      (input/output) REAL array, dimension (LDQ,N) */
00099 /*         On entry, the eigenvectors of the rank-1-perturbed matrix. */
00100 /*         On exit, the eigenvectors of the repaired tridiagonal matrix. */
00101 
00102 /*  LDQ    (input) INTEGER */
00103 /*         The leading dimension of the array Q.  LDQ >= max(1,N). */
00104 
00105 /*  INDXQ  (input/output) INTEGER array, dimension (N) */
00106 /*         On entry, the permutation which separately sorts the two */
00107 /*         subproblems in D into ascending order. */
00108 /*         On exit, the permutation which will reintegrate the */
00109 /*         subproblems back into sorted order, */
00110 /*         i.e. D( INDXQ( I = 1, N ) ) will be in ascending order. */
00111 
00112 /*  RHO    (input) REAL */
00113 /*         The subdiagonal entry used to create the rank-1 modification. */
00114 
00115 /*  CUTPNT (input) INTEGER */
00116 /*         The location of the last eigenvalue in the leading sub-matrix. */
00117 /*         min(1,N) <= CUTPNT <= N/2. */
00118 
00119 /*  WORK   (workspace) REAL array, dimension (4*N + N**2) */
00120 
00121 /*  IWORK  (workspace) INTEGER array, dimension (4*N) */
00122 
00123 /*  INFO   (output) INTEGER */
00124 /*          = 0:  successful exit. */
00125 /*          < 0:  if INFO = -i, the i-th argument had an illegal value. */
00126 /*          > 0:  if INFO = 1, an eigenvalue did not converge */
00127 
00128 /*  Further Details */
00129 /*  =============== */
00130 
00131 /*  Based on contributions by */
00132 /*     Jeff Rutter, Computer Science Division, University of California */
00133 /*     at Berkeley, USA */
00134 /*  Modified by Francoise Tisseur, University of Tennessee. */
00135 
00136 /*  ===================================================================== */
00137 
00138 /*     .. Local Scalars .. */
00139 /*     .. */
00140 /*     .. External Subroutines .. */
00141 /*     .. */
00142 /*     .. Intrinsic Functions .. */
00143 /*     .. */
00144 /*     .. Executable Statements .. */
00145 
00146 /*     Test the input parameters. */
00147 
00148     /* Parameter adjustments */
00149     --d__;
00150     q_dim1 = *ldq;
00151     q_offset = 1 + q_dim1;
00152     q -= q_offset;
00153     --indxq;
00154     --work;
00155     --iwork;
00156 
00157     /* Function Body */
00158     *info = 0;
00159 
00160     if (*n < 0) {
00161         *info = -1;
00162     } else if (*ldq < max(1,*n)) {
00163         *info = -4;
00164     } else /* if(complicated condition) */ {
00165 /* Computing MIN */
00166         i__1 = 1, i__2 = *n / 2;
00167         if (min(i__1,i__2) > *cutpnt || *n / 2 < *cutpnt) {
00168             *info = -7;
00169         }
00170     }
00171     if (*info != 0) {
00172         i__1 = -(*info);
00173         xerbla_("SLAED1", &i__1);
00174         return 0;
00175     }
00176 
00177 /*     Quick return if possible */
00178 
00179     if (*n == 0) {
00180         return 0;
00181     }
00182 
00183 /*     The following values are integer pointers which indicate */
00184 /*     the portion of the workspace */
00185 /*     used by a particular array in SLAED2 and SLAED3. */
00186 
00187     iz = 1;
00188     idlmda = iz + *n;
00189     iw = idlmda + *n;
00190     iq2 = iw + *n;
00191 
00192     indx = 1;
00193     indxc = indx + *n;
00194     coltyp = indxc + *n;
00195     indxp = coltyp + *n;
00196 
00197 
00198 /*     Form the z-vector which consists of the last row of Q_1 and the */
00199 /*     first row of Q_2. */
00200 
00201     scopy_(cutpnt, &q[*cutpnt + q_dim1], ldq, &work[iz], &c__1);
00202     cpp1 = *cutpnt + 1;
00203     i__1 = *n - *cutpnt;
00204     scopy_(&i__1, &q[cpp1 + cpp1 * q_dim1], ldq, &work[iz + *cutpnt], &c__1);
00205 
00206 /*     Deflate eigenvalues. */
00207 
00208     slaed2_(&k, n, cutpnt, &d__[1], &q[q_offset], ldq, &indxq[1], rho, &work[
00209             iz], &work[idlmda], &work[iw], &work[iq2], &iwork[indx], &iwork[
00210             indxc], &iwork[indxp], &iwork[coltyp], info);
00211 
00212     if (*info != 0) {
00213         goto L20;
00214     }
00215 
00216 /*     Solve Secular Equation. */
00217 
00218     if (k != 0) {
00219         is = (iwork[coltyp] + iwork[coltyp + 1]) * *cutpnt + (iwork[coltyp + 
00220                 1] + iwork[coltyp + 2]) * (*n - *cutpnt) + iq2;
00221         slaed3_(&k, n, cutpnt, &d__[1], &q[q_offset], ldq, rho, &work[idlmda], 
00222                  &work[iq2], &iwork[indxc], &iwork[coltyp], &work[iw], &work[
00223                 is], info);
00224         if (*info != 0) {
00225             goto L20;
00226         }
00227 
00228 /*     Prepare the INDXQ sorting permutation. */
00229 
00230         n1 = k;
00231         n2 = *n - k;
00232         slamrg_(&n1, &n2, &d__[1], &c__1, &c_n1, &indxq[1]);
00233     } else {
00234         i__1 = *n;
00235         for (i__ = 1; i__ <= i__1; ++i__) {
00236             indxq[i__] = i__;
00237 /* L10: */
00238         }
00239     }
00240 
00241 L20:
00242     return 0;
00243 
00244 /*     End of SLAED1 */
00245 
00246 } /* slaed1_ */


swiftnav
Author(s):
autogenerated on Sat Jun 8 2019 18:56:09