https://github.com/linbox-team/fflas-ffpack
Tip revision: a7801a65e9972b71558322e43812f5a7e08bbb4d authored by Clement Pernet on 14 November 2017, 16:52:10 UTC
fix parallel transpose
fix parallel transpose
Tip revision: a7801a6
test-fger.C
/* -*- mode: C++; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
// vim:sts=8:sw=8:ts=8:noet:sr:cino=>s,f0,{0,g0,(0,\:0,t0,+0,=s
/*
* Copyright (C) FFLAS-FFPACK
* Written by JG Dumas
* This file is Free Software and part of FFLAS-FFPACK.
*
* ========LICENCE========
* This file is part of the library FFLAS-FFPACK.
*
* FFLAS-FFPACK is free software: you can redistribute it and/or modify
* it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
* ========LICENCE========
*.
*/
//--------------------------------------------------------------------------
// Test for fger : 1 computation
//
//--------------------------------------------------------------------------
// Clement Pernet
//-------------------------------------------------------------------------
// #define __FFLASFFPACK_DEBUG
#define TIME 1
#include "fflas-ffpack/fflas-ffpack-config.h"
#include <iomanip>
#include <iostream>
#include <givaro/modular-int32.h>
#include <givaro/modular-balanced.h>
#include <givaro/givintprime.h>
#include "fflas-ffpack/utils/timer.h"
#include "fflas-ffpack/fflas/fflas.h"
#include "fflas-ffpack/utils/args-parser.h"
#include "fflas-ffpack/utils/test-utils.h"
#include "fflas-ffpack/utils/fflas_io.h"
using namespace std;
using namespace FFPACK;
using namespace FFLAS;
using Givaro::Modular;
using Givaro::ModularBalanced;
// checks that D = alpha . x . y^T + C
// WARNING
template<class Field>
bool check_fger(const Field & F,
const typename Field::Element_ptr Cd, // c0
const size_t m,
const size_t n,
const typename Field::Element & alpha,
const typename Field::Element_ptr x,
const size_t incx,
const typename Field::Element_ptr y,
const size_t incy,
const typename Field::Element_ptr C, // res
const size_t ldc
) {
bool wrong = false;
typedef typename Field::Element Element;
typedef typename Field::Element_ptr Element_ptr;
// std::cerr << "with(LinearAlgebra):" << std::endl;
// WriteMatrix(std::cerr <<"X:=",F, m, 1, x,incx, FflasMaple) << ';' << std::endl;
// WriteMatrix(std::cerr <<"Y:=Transpose(", F, n, 1, y, incy, FflasMaple) << ");" << std::endl;
// WriteMatrix(std::cerr <<"A:=",F, m, n, Cd, ldc, FflasMaple) << ';' << std::endl;
// F.write(std::cerr << "a:=", alpha) << ';' << std::endl;
// std::cerr << "q:=" << F.characteristic() << ';' << std::endl;
Element_ptr D = fflas_new (F,m,n);
fassign(F,m,n,Cd,n,D,n);
for(size_t i=0; i<m; ++i) {
Element tmp; F.init(tmp);
F.mul(tmp, alpha, *(x+i*incx) );
for(size_t j=0; j<n; j+=incy) {
F.axpyin(*(D+i*n+j), tmp, *(y+j) );
if ( !F.areEqual( *(D+i*n+j), *(C+i*ldc+j) ) ) {
wrong = true;
}
}
}
// WriteMatrix(std::cerr <<"d:=",F, m, n, D, n, FflasMaple) << ';' << std::endl;
// F.write(std::cerr, alpha) << "*X.Y+A,d;";
// F.write(std::cerr, alpha) << "*X.Y+A-d mod q;" << std::endl;
if ( wrong ){
size_t ici = 20 ;
std::cout<<"FAIL"<<std::endl;
std::cout << "a :" << alpha<<std::endl;
std::cout << "m :" << m << ", n : " << n << std::endl;
std::cout << "incx :" << incx << ", incy : " << incy << ", ldC : " << ldc << std::endl;
for (size_t i=0; i<m && ici; ++i){
for (size_t j =0; j<n && ici; ++j)
if (!F.areEqual( *(C+i*ldc+j), *(D+i*n+j) ) ) {
std::cout<<"Error C["<<i<<","<<j<<"]="
<<(*(C+i*ldc+j))<<" D["<<i<<","<<j<<"]="
<<(*(D+i*n+j))<<std::endl;
ici--;
}
}
if (m<80 && n<80) {
for (size_t i=0; i<m ; ++i){
for (size_t j =0; j<n ; ++j) {
if ( !F.areEqual( *(C+i*ldc+j), *(D+i*n+j) ) )
std::cout << 'X' ;
else
std::cout << '.' ;
}
std::cout << std::endl;
}
}
}
fflas_delete (D);
return !wrong ;
}
template<class Field, class RandIter>
bool launch_fger(const Field & F,
const size_t m,
const size_t n,
const typename Field::Element alpha,
const size_t ldc,
const size_t inca,
const size_t incb,
size_t iters,
RandIter& G)
{
bool ok = true;
typedef typename Field::Element_ptr Element_ptr;
Element_ptr A ;
FFLASFFPACK_check(inca >= 1);
Element_ptr B ;
FFLASFFPACK_check(incb >= 1);
Element_ptr C = fflas_new (F,m,ldc);
FFLASFFPACK_check(ldc >= n);
fzero(F,m,n,C,ldc);
Element_ptr D = fflas_new (F, m, n);
for(size_t i = 0;i<iters;++i){
A = fflas_new (F, m, inca);
RandomMatrix(F, m, inca, A, inca, G);
B = fflas_new (F, n, incb);
RandomMatrix(F, n, incb, B, incb, G);
RandomMatrix(F, m, n, C, ldc, G);
fassign(F,m,n,C,ldc,D,n);
fger (F,m,n,alpha, A, inca, B, incb, C,ldc);
ok = ok && check_fger(F, D, m,n,alpha, A, inca, B, incb, C,ldc);
fflas_delete(A);
fflas_delete(B);
if (!ok)
break;
}
fflas_delete (C);
fflas_delete (D);
return ok ;
}
template<class Field, class RandIter>
bool launch_fger_dispatch(const Field &F,
const size_t nn,
const typename Field::Element alpha,
const size_t iters,
RandIter& G)
{
bool ok = true;
size_t m,n;
size_t inca,incb,ldc;
//!@bug test for incx equal
//!@bug test for transpo
//!@todo does nbw actually do nbw recursive calls and then call blas (check ?) ?
// size_t ld = 13 ;
{
m = 1+(size_t)random()%nn;
n = 1+(size_t)random()%nn;
// lda = m+(size_t)random()%ld;
// ldb = 1+(size_t)random()%ld;
inca = 1;
incb = 1;
// ldc = n+(size_t)random()%ld;
ldc = n;
#ifdef __FFLASFFPACK_DEBUG
std::cout <<"q = "<<F.characteristic()<<" m,n = "<<m<<", "<<n<<" C := "
<<alpha<<".x * y^T + C";
#endif
ok = ok && launch_fger<Field>(F,m,n, alpha, ldc, inca, incb, iters, G);
#ifdef __FFLASFFPACK_DEBUG
std::cout<<(ok?" -> ok ":" -> KO")<<std::endl;
#endif
}
return ok ;
}
template <class Field>
bool run_with_field (int64_t q, uint64_t b, size_t n, size_t iters, uint64_t seed){
bool ok = true ;
int nbit=(int)iters;
while (ok && nbit){
typedef typename Field::Element Element ;
typedef typename Field::Element Element ;
Field* F= chooseField<Field>(q,b,seed);
if (F==NULL) return true;
std::ostringstream oss;
F->write(oss);
std::cout.fill('.');
std::cout<<"Checking ";
std::cout.width(45);
std::cout<<oss.str();
std::cout<<"... ";
typename Field::RandIter R(*F,0,seed++);
typename Field::NonZeroRandIter NZR(R);
//size_t k = 0 ;
//std::cout << k << "/24" << std::endl; ++k;
ok = ok && launch_fger_dispatch<Field>(*F,n,F->one,iters, R);
//std::cout << k << "/24" << std::endl; ++k;
ok = ok && launch_fger_dispatch<Field>(*F,n,F->zero,iters, R);
//std::cout << k << "/24" << std::endl; ++k;
ok = ok && launch_fger_dispatch<Field>(*F,n,F->mOne,iters, R);
//std::cout << k << "/24" << std::endl; ++k;
Element alpha ;
R.random(alpha);
ok = ok && launch_fger_dispatch<Field>(*F,n,alpha,iters, R);
if (!ok)
//std::cout << "\033[1;31mFAILED\033[0m "<<std::endl;
std::cout << "FAILED "<<std::endl;
else
//std::cout << "\033[1;32mPASSED\033[0m "<<std::endl;
std::cout << "PASSED "<<std::endl;
//std::cout<<std::endl;
nbit--;
delete F;
}
return ok;
}
int main(int argc, char** argv)
{
std::cout<<setprecision(17);
std::cerr<<setprecision(17);
size_t iters = 3 ;
long long q = -1 ;
uint64_t b = 0 ;
size_t n = 50 ;
bool loop = false;
uint64_t seed = getSeed();
Argument as[] = {
{ 'q', "-q Q", "Set the field characteristic (-1 for random).", TYPE_LONGLONG , &q },
{ 'b', "-b B", "Set the bitsize of the random characteristic.", TYPE_INT , &b },
{ 'n', "-n N", "Set the dimension of the matrix.", TYPE_INT , &n },
{ 'i', "-i R", "Set number of repetitions.", TYPE_INT , &iters },
{ 'l', "-loop Y/N", "run the test in an infinte loop.", TYPE_BOOL , &loop },
{ 's', "-s N", "Set the seed.", TYPE_UINT64 , &seed },
END_OF_ARGUMENTS
};
parseArguments(argc,argv,as);
bool ok = true;
do{
ok = ok && run_with_field<Modular<double> >(q,b,n,iters,seed);
ok = ok && run_with_field<ModularBalanced<double> >(q,b,n,iters,seed);
ok = ok && run_with_field<Modular<float> >(q,b,n,iters,seed);
ok = ok && run_with_field<ModularBalanced<float> >(q,b,n,iters,seed);
ok = ok && run_with_field<Modular<int32_t> >(q,b,n,iters,seed);
ok = ok && run_with_field<ModularBalanced<int32_t> >(q,b,n,iters,seed);
ok = ok && run_with_field<Modular<int64_t> >(q,b,n,iters,seed);
ok = ok && run_with_field<ModularBalanced<int64_t> >(q,b,n,iters,seed);
} while (loop && ok);
return !ok ;
}