test-ftrmm.C File Reference

tests all variants of applyP, ftrmm, ftrsm and fgemm for varions m,k,n and leading dimensions combinations. More...

#include <cassert>
#include "linbox/field/modular.h"
#include "linbox/linbox-config.h"
#include "fflas-ffpack/fflas/fflas.h"
#include "fflas-ffpack/ffpack/ffpack.h"
#include "linbox/integer.h"
#include "linbox/field/modular-balanced-int32.h"
#include "linbox/field/modular-int32.h"
#include "linbox/field/modular-balanced-double.h"
#include "linbox/field/modular-double.h"
#include "linbox/field/modular-balanced-float.h"
#include "linbox/field/modular-float.h"
#include "fflas-ffpack/utils/Matio.h"
#include "test-common.h"

Defines
#define	_LB_MAX_SZ   50
Functions
template<class Field , FFLAS::FFLAS_SIDE Side, FFLAS::FFLAS_UPLO UpLo, FFLAS::FFLAS_TRANSPOSE Trans, FFLAS::FFLAS_DIAG Diag>
int	test_ftrmm (std::ostream &report, const Field &F)
template<class Field >
int	test_applyP (std::ostream &report, const Field &F)
template<class Field , FFLAS::FFLAS_TRANSPOSE ATRANS, FFLAS::FFLAS_TRANSPOSE BTRANS>
int	test_fgemm (std::ostream &report, const Field &F)
int	main (int ac, char **av)

---

Detailed Description

tests all variants of applyP, ftrmm, ftrsm and fgemm for varions m,k,n and leading dimensions combinations.

Bug:

ftrmm fails on non-double transposed versions.

is fgemm correct on null A and B ?

Test:

FFLAS::ftrmm, FFPACK::applyP, FFLAS::ftrsm, FFLAS::fgemm

---

Define Documentation

#define _LB_MAX_SZ   50

Todo:

!

---

Function Documentation

int test_ftrmm	(	std::ostream &	report,
		const Field &	F
	)

Todo:

F.isInvertible()

Todo:

InvertibleRandomIter

Todo:

check ftrsm fails nicely with non invertible A !

int test_applyP	(	std::ostream &	report,
		const Field &	F
	)

Todo:

test NULL permutation

int test_fgemm	(	std::ostream &	report,
		const Field &	F
	)

Bug:

if beta != 0 but A and B are 0 (or don't exist) then C != beta C

Bug:

needs p prime.

int main	(	int	ac,
		char **	av
	)

Bug:

: this one completely fails :