summaryrefslogtreecommitdiff
path: root/numlib/ludecomp.c
diff options
context:
space:
mode:
authorJörg Frings-Fürst <debian@jff-webhosting.net>2014-09-01 13:56:46 +0200
committerJörg Frings-Fürst <debian@jff-webhosting.net>2014-09-01 13:56:46 +0200
commit22f703cab05b7cd368f4de9e03991b7664dc5022 (patch)
tree6f4d50beaa42328e24b1c6b56b6ec059e4ef21a5 /numlib/ludecomp.c
Initial import of argyll version 1.5.1-8debian/1.5.1-8
Diffstat (limited to 'numlib/ludecomp.c')
-rw-r--r--numlib/ludecomp.c500
1 files changed, 500 insertions, 0 deletions
diff --git a/numlib/ludecomp.c b/numlib/ludecomp.c
new file mode 100644
index 0000000..72e014a
--- /dev/null
+++ b/numlib/ludecomp.c
@@ -0,0 +1,500 @@
+/***************************************************/
+/* Linear Simultaeous equation solver */
+/***************************************************/
+
+/* General simultaneous equation solver. */
+/* Code was inspired by the algorithm decsribed in section 2.3 */
+/* of "Numerical Recipes in C", by W.H.Press, B.P.Flannery, */
+/* S.A.Teukolsky & W.T.Vetterling. */
+
+/*
+ * Copyright 2000 Graeme W. Gill
+ * All rights reserved.
+ *
+ * This material is licenced under the GNU AFFERO GENERAL PUBLIC LICENSE Version 3 :-
+ * see the License.txt file for licencing details.
+ */
+
+#include "numsup.h"
+#include "ludecomp.h"
+
+#undef DO_POLISH
+#undef DO_CHECK
+
+/* Solve the simultaneous linear equations A.X = B */
+/* Return 1 if the matrix is singular, 0 if OK */
+int
+solve_se(
+double **a, /* A[][] input matrix, returns LU decomposition of A */
+double *b, /* B[] input array, returns solution X[] */
+int n /* Dimensionality */
+) {
+ double rip; /* Row interchange parity */
+ int *pivx, PIVX[10];
+#if defined(DO_POLISH) || defined(DO_CHECK)
+ double **sa; /* save input matrix values */
+ double *sb; /* save input vector values */
+ int i, j;
+#endif
+
+ if (n <= 10)
+ pivx = PIVX;
+ else
+ pivx = ivector(0, n-1);
+
+#if defined(DO_POLISH) || defined(DO_CHECK)
+ sa = dmatrix(0, n-1, 0, n-1);
+ sb = dvector(0, n-1);
+
+ /* Copy input matrix and vector values */
+ for (i = 0; i < n; i++) {
+ sb[i] = b[i];
+ for (j = 0; j < n; j++)
+ sa[i][j] = a[i][j];
+ }
+#endif
+
+ if (lu_decomp(a, n, pivx, &rip)) {
+#if defined(DO_POLISH) || defined(DO_CHECK)
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+#endif
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 1;
+ }
+
+ lu_backsub(a, n, pivx, b);
+
+#ifdef DO_POLISH
+ lu_polish(sa, a, n, sb, b, pivx); /* Improve the solution */
+#endif
+
+#ifdef DO_CHECK
+ /* Check that the solution is correct */
+ for (i = 0; i < n; i++) {
+ double sum, temp;
+ sum = 0.0;
+ for (j = 0; j < n; j++)
+ sum += sa[i][j] * b[j];
+ temp = fabs(sum - sb[i]);
+ if (temp > 1e-6) {
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 2;
+ }
+ }
+#endif
+#if defined(DO_POLISH) || defined(DO_CHECK)
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+#endif
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 0;
+}
+
+/* Solve the simultaneous linear equations A.X = B, with polishing */
+/* Return 1 if the matrix is singular, 0 if OK */
+int
+polished_solve_se(
+double **a, /* A[][] input matrix, returns LU decomposition of A */
+double *b, /* B[] input array, returns solution X[] */
+int n /* Dimensionality */
+) {
+ double rip; /* Row interchange parity */
+ int *pivx, PIVX[10];
+ double **sa; /* save input matrix values */
+ double *sb; /* save input vector values */
+ int i, j;
+
+ if (n <= 10)
+ pivx = PIVX;
+ else
+ pivx = ivector(0, n-1);
+
+ sa = dmatrix(0, n-1, 0, n-1);
+ sb = dvector(0, n-1);
+
+ /* Copy source input matrix and vector values */
+ for (i = 0; i < n; i++) {
+ sb[i] = b[i];
+ for (j = 0; j < n; j++)
+ sa[i][j] = a[i][j];
+ }
+
+ if (lu_decomp(a, n, pivx, &rip)) {
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 1;
+ }
+
+ lu_backsub(a, n, pivx, b);
+
+ lu_polish(sa, a, n, sb, b, pivx); /* Improve the solution */
+
+#ifdef DO_CHECK
+ /* Check that the solution is correct */
+ for (i = 0; i < n; i++) {
+ double sum, temp;
+ sum = 0.0;
+ for (j = 0; j < n; j++)
+ sum += sa[i][j] * b[j];
+ temp = fabs(sum - sb[i]);
+ if (temp > 1e-6) {
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 2;
+ }
+ }
+#endif
+ free_dvector(sb, 0, n-1);
+ free_dmatrix(sa, 0, n-1, 0, n-1);
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 0;
+}
+
+/* Decompose the square matrix A[][] into lower and upper triangles */
+/* Return 1 if the matrix is singular. */
+int
+lu_decomp(
+double **a, /* A input array, output upper and lower triangles. */
+int n, /* Dimensionality */
+int *pivx, /* Return pivoting row permutations record */
+double *rip /* Row interchange parity, +/- 1.0, used for determinant */
+) {
+ int i, j;
+ double *rscale, RSCALE[10]; /* Implicit scaling of each row */
+
+ if (n <= 10)
+ rscale = RSCALE;
+ else
+ rscale = dvector(0, n-1);
+
+ /* For each row */
+ for (i = 0; i < n; i++) {
+ double big;
+ /* For each column in row */
+ for (big = 0.0, j=0; j < n; j++) {
+ double temp;
+ temp = fabs(a[i][j]);
+ if (temp > big)
+ big = temp;
+ }
+ if (fabs(big) <= DBL_MIN) {
+ if (rscale != RSCALE)
+ free_dvector(rscale, 0, n-1);
+ return 1; /* singular matrix */
+ }
+ rscale[i] = 1.0/big; /* Save the scaling */
+ }
+
+ /* For each column (Crout's method) */
+ for (*rip = 1.0, j = 0; j < n; j++) {
+ double big;
+ int k, bigi = 0;
+
+ /* For each row */
+ for (i = 0; i < j; i++) {
+ double sum;
+ sum = a[i][j];
+ for (k = 0; k < i; k++)
+ sum -= a[i][k] * a[k][j];
+ a[i][j] = sum;
+ }
+
+ /* Find largest pivot element */
+ for (big = 0.0, i = j; i < n; i++) {
+ double sum, temp;
+
+ sum = a[i][j];
+ for (k = 0; k < j; k++)
+ sum -= a[i][k] * a[k][j];
+ a[i][j] = sum;
+
+ temp = rscale[i] * fabs(sum); /* Figure of merit */
+ if (temp >= big) {
+ big = temp; /* Best so far */
+ bigi = i; /* Remember index */
+ }
+ }
+
+ /* If we need to interchange rows */
+ if (j != bigi) {
+ { /* Take advantage of matrix storage to swap pointers to rows */
+ double *temp;
+ temp = a[bigi];
+ a[bigi] = a[j];
+ a[j] = temp;
+ }
+ *rip = -(*rip); /* Another row interchange */
+ rscale[bigi] = rscale[j]; /* Interchange scale factor */
+ }
+
+ pivx[j] = bigi; /* Record pivot */
+ if (fabs(a[j][j]) <= DBL_MIN) {
+ if (rscale != RSCALE)
+ free_dvector(rscale, 0, n-1);
+ return 1; /* Pivot element is zero, so matrix is singular */
+ }
+
+ /* Divide by the pivot element */
+ if (j != (n-1)) {
+ double temp;
+ temp = 1.0/a[j][j];
+ for (i = j+1; i < n; i++)
+ a[i][j] *= temp;
+ }
+ }
+ if (rscale != RSCALE)
+ free_dvector(rscale, 0, n-1);
+ return 0;
+}
+
+/* Solve a set of simultaneous equations from the */
+/* LU decomposition, by back substitution. */
+void
+lu_backsub(
+double **a, /* A[][] LU decomposed matrix */
+int n, /* Dimensionality */
+int *pivx, /* Pivoting row permutations record */
+double *b /* Input B[] vecor, return X[] */
+) {
+ int i, j;
+ int nvi; /* When >= 0, indicates non-vanishing B[] index */
+
+ /* Forward substitution, undo pivoting on the fly */
+ for (nvi = -1, i = 0; i < n; i++) {
+ int px;
+ double sum;
+
+ px = pivx[i];
+ sum = b[px];
+ b[px] = b[i];
+ if (nvi >= 0) {
+ for (j = nvi; j < i; j++)
+ sum -= a[i][j] * b[j];
+ } else {
+ if (sum != 0.0)
+ nvi = i; /* Found non-vanishing element */
+ }
+ b[i] = sum;
+ }
+
+ /* Back substitution */
+ for (i = (n-1); i >= 0; i--) {
+ double sum;
+ sum = b[i];
+ for (j = i+1; j < n; j++)
+ sum -= a[i][j] * b[j];
+ b[i] = sum/a[i][i];
+ }
+}
+
+
+/* Improve a solution of equations */
+void
+lu_polish(
+double **a, /* Original A[][] matrix */
+double **lua, /* LU decomposition of A[][] */
+int n, /* Dimensionality */
+double *b, /* B[] vector of equation */
+double *x, /* X[] solution to be polished */
+int *pivx /* Pivoting row permutations record */
+) {
+ int i, j;
+ double *r, R[10]; /* Residuals */
+
+ if (n <= 10)
+ r = R;
+ else
+ r = dvector(0, n-1);
+
+ /* Accumulate the residual error */
+ for (i = 0; i < n; i++) {
+ double sum;
+ sum = -b[i];
+ for (j = 0; j < n; j++)
+ sum += a[i][j] * x[j];
+ r[i] = sum;
+ }
+
+ /* Solve for the error */
+ lu_backsub(lua, n, pivx, r);
+
+ /* Subtract error from the old solution */
+ for (i = 0; i < n; i++)
+ x[i] -= r[i];
+
+ if (r != R)
+ free_dvector(r, 0, n-1);
+}
+
+
+/* Invert a matrix A using lu decomposition */
+/* Return 1 if the matrix is singular, 0 if OK */
+int
+lu_invert(
+double **a, /* A[][] input matrix, returns inversion of A */
+int n /* Dimensionality */
+) {
+ int i, j;
+ double rip; /* Row interchange parity */
+ int *pivx, PIVX[10];
+ double **y;
+
+ if (n <= 10)
+ pivx = PIVX;
+ else
+ pivx = ivector(0, n-1);
+
+ if (lu_decomp(a, n, pivx, &rip)) {
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+ return 1;
+ }
+
+ /* Copy lu decomp. to y[][] */
+ y = dmatrix(0, n-1, 0, n-1);
+ for (i = 0; i < n; i++) {
+ for (j = 0; j < n; j++) {
+ y[i][j] = a[i][j];
+ }
+ }
+
+ /* Find inverse by columns */
+ for (i = 0; i < n; i++) {
+ for (j = 0; j < n; j++)
+ a[i][j] = 0.0;
+ a[i][i] = 1.0;
+ lu_backsub(y, n, pivx, a[i]);
+ }
+
+ /* Clean up */
+ free_dmatrix(y, 0, n-1, 0, n-1);
+ if (pivx != PIVX)
+ free_ivector(pivx, 0, n-1);
+
+ return 0;
+}
+
+#ifdef NEVER /* It's not clear that this is correct */
+int
+lu_polished_invert(
+double **a, /* A[][] input matrix, returns inversion of A */
+int n /* Dimensionality */
+) {
+ int i, j, k;
+ double **aa; /* saved a */
+ double **t1, **t2;
+
+ aa = dmatrix(0, n-1, 0, n-1);
+ t1 = dmatrix(0, n-1, 0, n-1);
+ t2 = dmatrix(0, n-1, 0, n-1);
+
+ /* Copy a to aa */
+ for (i = 0; i < n; i++) {
+ for (j = 0; j < n; j++)
+ aa[i][j] = a[i][j];
+ }
+
+ /* Invert a */
+ if ((i = lu_invert(a, n)) != 0) {
+ free_dmatrix(aa, 0, n-1, 0, n-1);
+ free_dmatrix(t1, 0, n-1, 0, n-1);
+ free_dmatrix(t2, 0, n-1, 0, n-1);
+ return i;
+ }
+
+ for (k = 0; k < 10; k++) {
+ matrix_mult(t1, n, n, aa, n, n, a, n, n);
+ for (i = 0; i < n; i++) {
+ for (j = 0; j < n; j++) {
+ t2[i][j] = a[i][j];
+ if (i == j)
+ t1[i][j] = 2.0 - t1[i][j];
+ else
+ t1[i][j] = 0.0 - t1[i][j];
+ }
+ }
+ matrix_mult(a, n, n, t2, n, n, t1, n, n);
+ }
+
+ free_dmatrix(aa, 0, n-1, 0, n-1);
+ free_dmatrix(t1, 0, n-1, 0, n-1);
+ free_dmatrix(t2, 0, n-1, 0, n-1);
+ return 0;
+}
+#endif
+
+/* Pseudo-Invert matrix A using lu decomposition */
+/* Return 1 if the matrix is singular, 0 if OK */
+int
+lu_psinvert(
+double **out, /* Output[0..N-1][0..M-1] */
+double **in, /* Input[0..M-1][0..N-1] input matrix */
+int m, /* Rows */
+int n /* Columns */
+) {
+ int rv = 0;
+ double **tr; /* Transpose */
+ double **sq; /* Square matrix */
+
+ tr = dmatrix(0, n-1, 0, m-1);
+ matrix_trans(tr, in, m, n);
+
+ /* Use left inverse */
+ if (m > n) {
+ sq = dmatrix(0, n-1, 0, n-1);
+
+ /* Multiply transpose by input */
+ if ((rv = matrix_mult(sq, n, n, tr, n, m, in, m, n)) == 0) {
+
+ /* Invert the square matrix */
+ if ((rv = lu_invert(sq, n)) == 0) {
+
+ /* Multiply inverted square by transpose */
+ rv = matrix_mult(out, n, m, sq, n, n, tr, n, m);
+ }
+ }
+ free_dmatrix(sq, 0, n-1, 0, n-1);
+
+ /* Use right inverse */
+ } else {
+ sq = dmatrix(0, m-1, 0, m-1);
+
+ /* Multiply input by transpose */
+ if ((rv = matrix_mult(sq, m, m, in, m, n, tr, n, m)) == 0) {
+
+ /* Invert the square matrix */
+ if ((rv = lu_invert(sq, m)) == 0) {
+
+ /* Multiply transpose by inverted square */
+ rv = matrix_mult(out, n, m, tr, n, m, sq, m, m);
+ }
+ }
+ free_dmatrix(sq, 0, m-1, 0, m-1);
+ }
+
+ free_dmatrix(tr, 0, n-1, 0, m-1);
+
+ return rv;
+}
+
+
+
+
+
+
+
+
+
+
+