📄 amatrix.cpp
字号:
/*M///////////////////////////////////////////////////////////////////////////////////////
//
// IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
//
// By downloading, copying, installing or using the software you agree to this license.
// If you do not agree to this license, do not download, install,
// copy or use the software.
//
//
// Intel License Agreement
// For Open Source Computer Vision Library
//
// Copyright (C) 2000, Intel Corporation, all rights reserved.
// Third party copyrights are property of their respective owners.
//
// Redistribution and use in source and binary forms, with or without modification,
// are permitted provided that the following conditions are met:
//
// * Redistribution's of source code must retain the above copyright notice,
// this list of conditions and the following disclaimer.
//
// * Redistribution's in binary form must reproduce the above copyright notice,
// this list of conditions and the following disclaimer in the documentation
// and/or other materials provided with the distribution.
//
// * The name of Intel Corporation may not be used to endorse or promote products
// derived from this software without specific prior written permission.
//
// This software is provided by the copyright holders and contributors "as is" and
// any express or implied warranties, including, but not limited to, the implied
// warranties of merchantability and fitness for a particular purpose are disclaimed.
// In no event shall the Intel Corporation or contributors be liable for any direct,
// indirect, incidental, special, exemplary, or consequential damages
// (including, but not limited to, procurement of substitute goods or services;
// loss of use, data, or profits; or business interruption) however caused
// and on any theory of liability, whether in contract, strict liability,
// or tort (including negligence or otherwise) arising in any way out of
// the use of this software, even if advised of the possibility of such damage.
//
//M*/
#define ADD 0
#define SUB 1
#define SCALE 2
#define MULT 3
#define TRANSP 4
#define INVERT 5
#define MAHAL 6
#include "CvmTest.h"
#include <float.h>
#define PI 3.1415f
#define FEPS 0.001f
/* Testing parameters */
static char* func_names[] =
{
"cvmAdd",
"cvmSub",
"cvmScale",
"cvmMul",
"cvmTranspose",
"cvmInvert",
"cvmMahalanobis"
};
static char* test_desc = "Matrix - Matrix operations";
static int length;
static int step;
static int read_param = 0;
static int test_type=6;
static int fcaMatrMatr( void* prm )
{
int i,j;
int Errors = 0;
long lParam = (long)prm;
AtsRandState state;
static int read_param = 0;
/* Initialization global parameters */
if( !read_param )
{
/* Determine which tests are needed to run */
trsCaseRead( &test_type, "/A/S/SC/M/T/I/a", "a",
"a - all, A - add, S - sub, SC - scale, M - multiply, T-Transpose,I - Invert" );
/* Reading test-parameters */
trsiRead( &length, "16", " size of matrices" );
read_param = 1;
}
if((test_type!=6)&&(test_type!=lParam)) return TRS_UNDEF;
CvMat flSrcV1 = cvMat(length,1,CV_MAT32F,NULL);
CvMat flSrcV2 = cvMat(length,1,CV_MAT32F,NULL);
CvMat flsrc1 = cvMat(length, length,CV_MAT32F,NULL);
CvMat flsrc2 = cvMat(length, length,CV_MAT32F,NULL);
CvMat fldst = cvMat(length, length,CV_MAT32F,NULL);
CvMat flmklDest = cvMat(length, length,CV_MAT32F,NULL);
CvMat flmklZero = cvMat(length, length,CV_MAT32F,NULL);
CvMat flmklUnit = cvMat(length, length,CV_MAT32F,NULL);
CvMat dbsrc1 = cvMat(length, length,CV_MAT64D,NULL);
CvMat dbsrc2 = cvMat(length, length,CV_MAT64D,NULL);
CvMat dbdst = cvMat(length, length,CV_MAT64D,NULL);
CvMat dbmklDest = cvMat(length, length,CV_MAT64D,NULL);
CvMat dbmklZero = cvMat(length, length,CV_MAT64D,NULL);
CvMat dbmklUnit = cvMat(length, length,CV_MAT64D,NULL);
cvmAlloc(&flSrcV1);
cvmAlloc(&flSrcV2);
cvmAlloc(&flsrc1);
cvmAlloc(&flsrc2);
cvmAlloc(&fldst);
cvmAlloc(&flmklZero);
cvmAlloc(&flmklUnit);
cvmAlloc(&flmklDest);
cvmAlloc(&dbsrc1);
cvmAlloc(&dbsrc2);
cvmAlloc(&dbdst);
cvmAlloc(&dbmklZero);
cvmAlloc(&dbmklUnit);
cvmAlloc(&dbmklDest);
atsRandInit( &state, -256, 256, 10 );
atsbRand32f( &state, flSrcV1.data.fl, length );
atsbRand32f( &state, flSrcV2.data.fl, length );
atsbRand32f( &state, flsrc1.data.fl, length*length );
atsbRand32f( &state, flsrc2.data.fl, length*length );
atsbRand64d( &state, dbsrc1.data.db, length*length );
atsbRand64d( &state, dbsrc2.data.db, length*length );
cvmSetZero(&flmklZero);
cvmSetIdentity(&flmklUnit);
cvmSetZero(&dbmklZero);
cvmSetIdentity(&dbmklUnit);
switch( lParam )
{
case ADD:
cvmAdd( &flsrc1,&flsrc2,&fldst);
cvmCopy(&flsrc2,&flmklDest);
/* mkl check */
cblas_sgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,flsrc1.data.fl,length,
flmklUnit.data.fl,length,1,flmklDest.data.fl,length);
/* check */
Errors += atsCompSinglePrec( flmklDest.data.fl ,fldst.data.fl, length*length, FEPS );
cvmAdd(&dbsrc1,&dbsrc2,&dbdst);
cvmCopy(&dbsrc2,&dbmklDest);
/* mkl check */
cblas_dgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,dbsrc1.data.db,length,
dbmklUnit.data.db,length,1,dbmklDest.data.db,length);
/* check */
Errors += atsCompDoublePrec( dbmklDest.data.db ,dbdst.data.db, length*length, FEPS );
break;
case SUB:
cvmSub( &flsrc1,&flsrc2,&fldst);
cvmCopy(&flsrc2,&flmklDest);
/* mkl check */
cblas_sgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,flsrc1.data.fl,length,
flmklUnit.data.fl,length,-1,flmklDest.data.fl,length);
/* check */
Errors += atsCompSinglePrec( flmklDest.data.fl ,fldst.data.fl, length*length, FEPS );
cvmSub(&dbsrc1,&dbsrc2,&dbdst);
cvmCopy(&dbsrc2,&dbmklDest);
/* mkl check */
cblas_dgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,dbsrc1.data.db,length,
dbmklUnit.data.db,length,-1,dbmklDest.data.db,length);
/* check */
Errors += atsCompDoublePrec( dbmklDest.data.db ,dbdst.data.db, length*length, FEPS );
break;
case SCALE:
cvmScale(&flsrc2,&fldst,PI);
cvmCopy(&flsrc2,&flmklDest);
/* mkl check */
cblas_sgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,flmklZero.data.fl,length,
flmklZero.data.fl,length,PI,flmklDest.data.fl,length);
/* check */
Errors += atsCompSinglePrec( flmklDest.data.fl ,fldst.data.fl, length*length, FEPS );
cvmScale(&dbsrc2,&dbdst,PI);
cvmCopy(&dbsrc2,&dbmklDest);
/* mkl check */
cblas_dgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1,dbmklZero.data.db,length,
dbmklZero.data.db,length,PI,dbmklDest.data.db,length);
/* check */
Errors += atsCompDoublePrec( dbmklDest.data.db ,dbdst.data.db, length*length, FEPS );
break;
case MULT:
cvmMul(&flsrc1,&flsrc2,&fldst) ;
/* mkl check */
cblas_sgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1, flsrc1.data.fl, length,
flsrc2.data.fl,length,0,flmklZero.data.fl,length);
Errors += atsCompSinglePrec( flmklZero.data.fl ,fldst.data.fl, length*length, FEPS );
cvmMul(&dbsrc1,&dbsrc2,&dbdst);
/* mkl check */
cblas_dgemm( CblasRowMajor, CblasNoTrans, CblasNoTrans,
length, length, length, 1, dbsrc1.data.db, length,
dbsrc2.data.db,length,0,dbmklZero.data.db,length);
Errors += atsCompDoublePrec( dbmklZero.data.db ,dbdst.data.db, length*length, FEPS );
break;
case TRANSP:
cvmTranspose(&flsrc1,&fldst);
/* get data to simple mode */
for(i = 0; i < length; i++)
for(j = 0; j < length; j++)
{
if(cvmGet(&flsrc1,i,j)!=cvmGet(&fldst,j,i))
Errors++;
}
cvmTranspose(&dbsrc1,&dbdst);
/* get data to simple mode */
for(i = 0; i < length; i++)
for(j = 0; j < length; j++)
{
if(cvmGet(&dbsrc1,i,j)!=cvmGet(&dbdst,j,i))
Errors++;
}
break;
case INVERT:
cvmInvert(&flsrc1,&flsrc2);
cvmMul(&flsrc1,&flsrc2,&fldst);
Errors += atsCompSinglePrec( flmklUnit.data.fl ,fldst.data.fl, length*length, FEPS );
cvmInvert(&dbsrc1,&dbsrc2);
cvmMul(&dbsrc1,&dbsrc2,&dbdst);
Errors += atsCompDoublePrec( dbmklUnit.data.db ,dbdst.data.db, length*length, FEPS );
break;
case MAHAL:
double testval = cvmMahalanobis(&flSrcV1,&flSrcV2,&flsrc1);
cblas_saxpy(length,-1,flSrcV2.data.fl,1,flSrcV1.data.fl,1);
cblas_sgemv( CblasRowMajor,CblasNoTrans,length, length, 1, flsrc1.data.fl,length,flSrcV1.data.fl, 1,0,
flSrcV2.data.fl,1);
double val = sqrt(cblas_sdot(length,flSrcV1.data.fl,1,flSrcV2.data.fl,1));
Errors += (int)fabs(val-testval);
}
cvmFree(&flSrcV1);
cvmFree(&flSrcV2);
cvmFree(&flsrc1);
cvmFree(&flsrc2);
cvmFree(&fldst);
cvmFree(&flmklZero);
cvmFree(&flmklUnit);
cvmFree(&flmklDest);
cvmFree(&dbsrc1);
cvmFree(&dbsrc2);
cvmFree(&dbdst);
cvmFree(&dbmklZero);
cvmFree(&dbmklUnit);
cvmFree(&dbmklDest);
if( Errors == 0 ) return trsResult( TRS_OK, "No errors fixed for this text" );
else return trsResult( TRS_FAIL, "Total fixed %d errors", Errors );
}
void InitAMatr( void )
{
/* Registering test functions */
trsRegArg( func_names[0], test_desc, atsAlgoClass, fcaMatrMatr, ADD );
trsRegArg( func_names[1], test_desc, atsAlgoClass, fcaMatrMatr, SUB );
trsRegArg( func_names[2], test_desc, atsAlgoClass, fcaMatrMatr, SCALE );
trsRegArg( func_names[3], test_desc, atsAlgoClass, fcaMatrMatr, MULT );
trsRegArg( func_names[4], test_desc, atsAlgoClass, fcaMatrMatr, TRANSP );
trsRegArg( func_names[5], test_desc, atsAlgoClass, fcaMatrMatr, INVERT );
trsRegArg( func_names[6], test_desc, atsAlgoClass, fcaMatrMatr, MAHAL );
} /* InitAMatrMatr */
/* End of File */
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -