Mercurial > hg > octave-nkf
diff src/OPERATORS/op-fcm-fcs.cc @ 7789:82be108cc558
First attempt at single precision tyeps
* * *
corrections to qrupdate single precision routines
* * *
prefer demotion to single over promotion to double
* * *
Add single precision support to log2 function
* * *
Trivial PROJECT file update
* * *
Cache optimized hermitian/transpose methods
* * *
Add tests for tranpose/hermitian and ChangeLog entry for new transpose code
author | David Bateman <dbateman@free.fr> |
---|---|
date | Sun, 27 Apr 2008 22:34:17 +0200 |
parents | |
children | 87865ed7405f |
line wrap: on
line diff
new file mode 100644 --- /dev/null +++ b/src/OPERATORS/op-fcm-fcs.cc @@ -0,0 +1,176 @@ +/* + +Copyright (C) 1996, 1997, 2000, 2001, 2002, 2003, 2004, 2005, 2006, + 2007 John W. Eaton + +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 "gripes.h" +#include "oct-obj.h" +#include "ov.h" +#include "ov-cx-mat.h" +#include "ov-flt-cx-mat.h" +#include "ov-flt-complex.h" +#include "ov-typeinfo.h" +#include "ops.h" +#include "xdiv.h" +#include "xpow.h" + +// complex matrix by complex scalar ops. + +DEFNDBINOP_OP (add, float_complex_matrix, float_complex, + float_complex_array, float_complex, +) +DEFNDBINOP_OP (sub, float_complex_matrix, float_complex, + float_complex_array, float_complex, -) +DEFNDBINOP_OP (mul, float_complex_matrix, float_complex, + float_complex_array, float_complex, *) + +DEFBINOP (div, float_complex_matrix, float_complex) +{ + CAST_BINOP_ARGS (const octave_float_complex_matrix&, + const octave_float_complex&); + + FloatComplex d = v2.float_complex_value (); + + if (d == static_cast<float>(0.0)) + gripe_divide_by_zero (); + + return octave_value (v1.float_complex_array_value () / d); +} + +DEFBINOP_FN (pow, float_complex_matrix, float_complex, xpow) + +DEFBINOP (ldiv, float_complex_matrix, float_complex) +{ + CAST_BINOP_ARGS (const octave_float_complex_matrix&, + const octave_float_complex&); + + FloatComplexMatrix m1 = v1.float_complex_matrix_value (); + FloatComplexMatrix m2 = v2.float_complex_matrix_value (); + MatrixType typ = v1.matrix_type (); + + FloatComplexMatrix ret = xleftdiv (m1, m2, typ); + v1.matrix_type (typ); + return ret; +} + +DEFNDBINOP_FN (lt, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_lt) +DEFNDBINOP_FN (le, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_le) +DEFNDBINOP_FN (eq, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_eq) +DEFNDBINOP_FN (ge, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_ge) +DEFNDBINOP_FN (gt, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_gt) +DEFNDBINOP_FN (ne, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_ne) + +DEFNDBINOP_OP (el_mul, float_complex_matrix, float_complex, + float_complex_array, float_complex, *) + +DEFBINOP (el_div, float_complex_matrix, float_complex) +{ + CAST_BINOP_ARGS (const octave_float_complex_matrix&, + const octave_float_complex&); + + FloatComplex d = v2.float_complex_value (); + + if (d == static_cast<float>(0.0)) + gripe_divide_by_zero (); + + return octave_value (v1.float_complex_array_value () / d); +} + +DEFNDBINOP_FN (el_pow, float_complex_matrix, float_complex, + float_complex_array, float_complex, elem_xpow) + +DEFBINOP (el_ldiv, float_complex_matrix, float_complex) +{ + CAST_BINOP_ARGS (const octave_float_complex_matrix&, + const octave_float_complex&); + + return x_el_div (v2.float_complex_value (), v1.float_complex_array_value ()); +} + +DEFNDBINOP_FN (el_and, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_and) +DEFNDBINOP_FN (el_or, float_complex_matrix, float_complex, + float_complex_array, float_complex, mx_el_or) + +DEFNDCATOP_FN (fcm_fcs, float_complex_matrix, float_complex, + float_complex_array, float_complex_array, concat) + +DEFNDASSIGNOP_FN (assign, float_complex_matrix, float_complex, + float_complex_array, assign) +DEFNDASSIGNOP_FN (dbl_assign, complex_matrix, float_complex, + complex_array, assign) + +void +install_fcm_fcs_ops (void) +{ + INSTALL_BINOP (op_add, octave_float_complex_matrix, + octave_float_complex, add); + INSTALL_BINOP (op_sub, octave_float_complex_matrix, + octave_float_complex, sub); + INSTALL_BINOP (op_mul, octave_float_complex_matrix, + octave_float_complex, mul); + INSTALL_BINOP (op_div, octave_float_complex_matrix, + octave_float_complex, div); + INSTALL_BINOP (op_pow, octave_float_complex_matrix, + octave_float_complex, pow); + INSTALL_BINOP (op_ldiv, octave_float_complex_matrix, + octave_float_complex, ldiv); + INSTALL_BINOP (op_lt, octave_float_complex_matrix, octave_float_complex, lt); + INSTALL_BINOP (op_le, octave_float_complex_matrix, octave_float_complex, le); + INSTALL_BINOP (op_eq, octave_float_complex_matrix, octave_float_complex, eq); + INSTALL_BINOP (op_ge, octave_float_complex_matrix, octave_float_complex, ge); + INSTALL_BINOP (op_gt, octave_float_complex_matrix, octave_float_complex, gt); + INSTALL_BINOP (op_ne, octave_float_complex_matrix, octave_float_complex, ne); + INSTALL_BINOP (op_el_mul, octave_float_complex_matrix, + octave_float_complex, el_mul); + INSTALL_BINOP (op_el_div, octave_float_complex_matrix, + octave_float_complex, el_div); + INSTALL_BINOP (op_el_pow, octave_float_complex_matrix, + octave_float_complex, el_pow); + INSTALL_BINOP (op_el_ldiv, octave_float_complex_matrix, + octave_float_complex, el_ldiv); + INSTALL_BINOP (op_el_and, octave_float_complex_matrix, + octave_float_complex, el_and); + INSTALL_BINOP (op_el_or, octave_float_complex_matrix, + octave_float_complex, el_or); + + INSTALL_CATOP (octave_float_complex_matrix, octave_float_complex, fcm_fcs); + + INSTALL_ASSIGNOP (op_asn_eq, octave_float_complex_matrix, + octave_float_complex, assign); + INSTALL_ASSIGNOP (op_asn_eq, octave_complex_matrix, + octave_float_complex, dbl_assign); +} + +/* +;;; Local Variables: *** +;;; mode: C++ *** +;;; End: *** +*/