/**
* Generalized Least Squares via *GELSD.
*
* <p>Note that B must be padded to contain the solution matrix. This occurs when A has fewer rows
* than columns.
*
* <p>For example: in A * X = B, A is (m,n), X is (n,k) and B is (m,k). Now if m < n, since B is
* overwritten to contain the solution (in classical LAPACK style), B needs to be padded to be an
* (n,k) matrix.
*
* <p>Likewise, if m > n, the solution consists only of the first n rows of B.
*
* @param A an (m,n) matrix
* @param B an (max(m,n), k) matrix (well, at least)
*/
public static void gelsd(FloatMatrix A, FloatMatrix B) {
int m = A.rows;
int n = A.columns;
int nrhs = B.columns;
int minmn = min(m, n);
int maxmn = max(m, n);
if (B.rows < maxmn) {
throw new SizeException("Result matrix B must be padded to contain the solution matrix X!");
}
int smlsiz = NativeBlas.ilaenv(9, "DGELSD", "", m, n, nrhs, 0);
int nlvl = max(0, (int) log2(minmn / (smlsiz + 1)) + 1);
// System.err.printf("GELSD\n");
// System.err.printf("m = %d, n = %d, nrhs = %d\n", m, n, nrhs);
// System.err.printf("smlsiz = %d, nlvl = %d\n", smlsiz, nlvl);
// System.err.printf("iwork size = %d\n", 3 * minmn * nlvl + 11 * minmn);
int[] iwork = new int[3 * minmn * nlvl + 11 * minmn];
float[] s = new float[minmn];
int[] rank = new int[1];
int info =
NativeBlas.sgelsd(m, n, nrhs, A.data, 0, m, B.data, 0, B.rows, s, 0, -1, rank, 0, iwork, 0);
if (info == 0) {
return;
} else if (info < 0) {
throw new LapackArgumentException("DGESD", -info);
} else if (info > 0) {
throw new LapackConvergenceException(
"DGESD",
info + " off-diagonal elements of an intermediat bidiagonal form did not converge to 0.");
}
}
