view libinterp/corefcn/rcond.cc @ 20939:b17fda023ca6

maint: Use new C++ archetype in more files. Place input validation first in files. Move declaration of retval down in function to be closer to point of usage. Eliminate else clause after if () error. Use "return ovl()" where it makes sense. * find.cc, gammainc.cc, gcd.cc, getgrent.cc, getpwent.cc, givens.cc, graphics.cc, help.cc, hess.cc, hex2num.cc, input.cc, kron.cc, load-path.cc, load-save.cc, lookup.cc, mappers.cc, matrix_type.cc, mgorth.cc, nproc.cc, ordschur.cc, pager.cc, pinv.cc, pr-output.cc, profiler.cc, psi.cc, quad.cc, rcond.cc, regexp.cc, schur.cc, sighandlers.cc, sparse.cc, str2double.cc, strfind.cc, strfns.cc, sub2ind.cc, svd.cc, sylvester.cc, symtab.cc, syscalls.cc, sysdep.cc, time.cc, toplev.cc, tril.cc, tsearch.cc, typecast.cc, urlwrite.cc, utils.cc, variables.cc, __delaunayn__.cc, __eigs__.cc, __glpk__.cc, __magick_read__.cc, __osmesa_print__.cc, __voronoi__.cc, amd.cc, audiodevinfo.cc, audioread.cc, chol.cc, colamd.cc, dmperm.cc, fftw.cc, qr.cc, symbfact.cc, symrcm.cc, ov-bool-mat.cc, ov-cell.cc, ov-class.cc, ov-classdef.cc, ov-fcn-handle.cc, ov-fcn-inline.cc, ov-flt-re-mat.cc, ov-java.cc, ov-null-mat.cc, ov-oncleanup.cc, ov-re-mat.cc, ov-struct.cc, ov-typeinfo.cc, ov-usr-fcn.cc, ov.cc, octave.cc: Use new C++ archetype in more files.
author Rik <rik@octave.org>
date Fri, 18 Dec 2015 15:37:22 -0800
parents 1142cf6abc0d
children 48b2ad5ee801
line wrap: on
line source

/*

Copyright (C) 2008-2015 David Bateman

This file is part of Octave.

Octave is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3 of the License, or (at your
option) any later version.

Octave is distributed in the hope that it will be useful, but WITHOUT
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
for more details.

You should have received a copy of the GNU General Public License
along with Octave; see the file COPYING.  If not, see
<http://www.gnu.org/licenses/>.

*/

#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include "defun.h"
#include "error.h"
#include "gripes.h"
#include "oct-obj.h"
#include "utils.h"

DEFUN (rcond, args, ,
       "-*- texinfo -*-\n\
@deftypefn {} {@var{c} =} rcond (@var{A})\n\
Compute the 1-norm estimate of the reciprocal condition number as returned\n\
by @sc{lapack}.\n\
\n\
If the matrix is well-conditioned then @var{c} will be near 1 and if the\n\
matrix is poorly conditioned it will be close to 0.\n\
\n\
The matrix @var{A} must not be sparse.  If the matrix is sparse then\n\
@code{condest (@var{A})} or @code{rcond (full (@var{A}))} should be used\n\
instead.\n\
@seealso{cond, condest}\n\
@end deftypefn")
{
  if (args.length () != 1)
    print_usage ();

  octave_value retval;

  if (args(0).is_sparse_type ())
    error ("rcond: for sparse matrices use 'rcond (full (a))' or 'condest (a)' instead");

  if (args(0).is_single_type ())
    {
      if (args(0).is_complex_type ())
        {
          FloatComplexMatrix m = args(0).float_complex_matrix_value ();
          MatrixType mattyp;
          retval = m.rcond (mattyp);
          args(0).matrix_type (mattyp);
        }
      else
        {
          FloatMatrix m = args(0).float_matrix_value ();
          MatrixType mattyp;
          retval = m.rcond (mattyp);
          args(0).matrix_type (mattyp);
        }
    }
  else if (args(0).is_complex_type ())
    {
      ComplexMatrix m = args(0).complex_matrix_value ();
      MatrixType mattyp;
      retval = m.rcond (mattyp);
      args(0).matrix_type (mattyp);
    }
  else
    {
      Matrix m = args(0).matrix_value ();
      MatrixType mattyp;
      retval = m.rcond (mattyp);
      args(0).matrix_type (mattyp);
    }

  return retval;
}

/*
%!assert (rcond (eye (2)), 1)
%!assert (rcond (ones (2)), 0)
%!assert (rcond ([1 1; 2 1]), 1/9)
%!assert (rcond (magic (4)), 0, eps)

%!shared x, sx
%! x = [-5.25, -2.25; -2.25, 1] * eps () + ones (2) / 2;
%! sx = [-5.25, -2.25; -2.25, 1] * eps ("single") + ones (2) / 2;
%!assert (rcond (x) < eps ());
%!assert (rcond (sx) < eps ('single'));
%!assert (rcond (x*i) < eps ());
%!assert (rcond (sx*i) < eps ('single'));

*/