dpoequ.c
Go to the documentation of this file.
00001 /* dpoequ.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 /* Subroutine */ int dpoequ_(integer *n, doublereal *a, integer *lda, 
00017         doublereal *s, doublereal *scond, doublereal *amax, integer *info)
00018 {
00019     /* System generated locals */
00020     integer a_dim1, a_offset, i__1;
00021     doublereal d__1, d__2;
00022 
00023     /* Builtin functions */
00024     double sqrt(doublereal);
00025 
00026     /* Local variables */
00027     integer i__;
00028     doublereal smin;
00029     extern /* Subroutine */ int xerbla_(char *, integer *);
00030 
00031 
00032 /*  -- LAPACK routine (version 3.2) -- */
00033 /*     Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. */
00034 /*     November 2006 */
00035 
00036 /*     .. Scalar Arguments .. */
00037 /*     .. */
00038 /*     .. Array Arguments .. */
00039 /*     .. */
00040 
00041 /*  Purpose */
00042 /*  ======= */
00043 
00044 /*  DPOEQU computes row and column scalings intended to equilibrate a */
00045 /*  symmetric positive definite matrix A and reduce its condition number */
00046 /*  (with respect to the two-norm).  S contains the scale factors, */
00047 /*  S(i) = 1/sqrt(A(i,i)), chosen so that the scaled matrix B with */
00048 /*  elements B(i,j) = S(i)*A(i,j)*S(j) has ones on the diagonal.  This */
00049 /*  choice of S puts the condition number of B within a factor N of the */
00050 /*  smallest possible condition number over all possible diagonal */
00051 /*  scalings. */
00052 
00053 /*  Arguments */
00054 /*  ========= */
00055 
00056 /*  N       (input) INTEGER */
00057 /*          The order of the matrix A.  N >= 0. */
00058 
00059 /*  A       (input) DOUBLE PRECISION array, dimension (LDA,N) */
00060 /*          The N-by-N symmetric positive definite matrix whose scaling */
00061 /*          factors are to be computed.  Only the diagonal elements of A */
00062 /*          are referenced. */
00063 
00064 /*  LDA     (input) INTEGER */
00065 /*          The leading dimension of the array A.  LDA >= max(1,N). */
00066 
00067 /*  S       (output) DOUBLE PRECISION array, dimension (N) */
00068 /*          If INFO = 0, S contains the scale factors for A. */
00069 
00070 /*  SCOND   (output) DOUBLE PRECISION */
00071 /*          If INFO = 0, S contains the ratio of the smallest S(i) to */
00072 /*          the largest S(i).  If SCOND >= 0.1 and AMAX is neither too */
00073 /*          large nor too small, it is not worth scaling by S. */
00074 
00075 /*  AMAX    (output) DOUBLE PRECISION */
00076 /*          Absolute value of largest matrix element.  If AMAX is very */
00077 /*          close to overflow or very close to underflow, the matrix */
00078 /*          should be scaled. */
00079 
00080 /*  INFO    (output) INTEGER */
00081 /*          = 0:  successful exit */
00082 /*          < 0:  if INFO = -i, the i-th argument had an illegal value */
00083 /*          > 0:  if INFO = i, the i-th diagonal element is nonpositive. */
00084 
00085 /*  ===================================================================== */
00086 
00087 /*     .. Parameters .. */
00088 /*     .. */
00089 /*     .. Local Scalars .. */
00090 /*     .. */
00091 /*     .. External Subroutines .. */
00092 /*     .. */
00093 /*     .. Intrinsic Functions .. */
00094 /*     .. */
00095 /*     .. Executable Statements .. */
00096 
00097 /*     Test the input parameters. */
00098 
00099     /* Parameter adjustments */
00100     a_dim1 = *lda;
00101     a_offset = 1 + a_dim1;
00102     a -= a_offset;
00103     --s;
00104 
00105     /* Function Body */
00106     *info = 0;
00107     if (*n < 0) {
00108         *info = -1;
00109     } else if (*lda < max(1,*n)) {
00110         *info = -3;
00111     }
00112     if (*info != 0) {
00113         i__1 = -(*info);
00114         xerbla_("DPOEQU", &i__1);
00115         return 0;
00116     }
00117 
00118 /*     Quick return if possible */
00119 
00120     if (*n == 0) {
00121         *scond = 1.;
00122         *amax = 0.;
00123         return 0;
00124     }
00125 
00126 /*     Find the minimum and maximum diagonal elements. */
00127 
00128     s[1] = a[a_dim1 + 1];
00129     smin = s[1];
00130     *amax = s[1];
00131     i__1 = *n;
00132     for (i__ = 2; i__ <= i__1; ++i__) {
00133         s[i__] = a[i__ + i__ * a_dim1];
00134 /* Computing MIN */
00135         d__1 = smin, d__2 = s[i__];
00136         smin = min(d__1,d__2);
00137 /* Computing MAX */
00138         d__1 = *amax, d__2 = s[i__];
00139         *amax = max(d__1,d__2);
00140 /* L10: */
00141     }
00142 
00143     if (smin <= 0.) {
00144 
00145 /*        Find the first non-positive diagonal element and return. */
00146 
00147         i__1 = *n;
00148         for (i__ = 1; i__ <= i__1; ++i__) {
00149             if (s[i__] <= 0.) {
00150                 *info = i__;
00151                 return 0;
00152             }
00153 /* L20: */
00154         }
00155     } else {
00156 
00157 /*        Set the scale factors to the reciprocals */
00158 /*        of the diagonal elements. */
00159 
00160         i__1 = *n;
00161         for (i__ = 1; i__ <= i__1; ++i__) {
00162             s[i__] = 1. / sqrt(s[i__]);
00163 /* L30: */
00164         }
00165 
00166 /*        Compute SCOND = min(S(I)) / max(S(I)) */
00167 
00168         *scond = sqrt(smin) / sqrt(*amax);
00169     }
00170     return 0;
00171 
00172 /*     End of DPOEQU */
00173 
00174 } /* dpoequ_ */


swiftnav
Author(s):
autogenerated on Sat Jun 8 2019 18:55:47