1/*M///////////////////////////////////////////////////////////////////////////////////////
2//
3//  IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING.
4//
5//  By downloading, copying, installing or using the software you agree to this license.
6//  If you do not agree to this license, do not download, install,
7//  copy or use the software.
8//
9//
10//                        Intel License Agreement
11//
12// Copyright (C) 2000, Intel Corporation, all rights reserved.
13// Third party copyrights are property of their respective owners.
14//
15// Redistribution and use in source and binary forms, with or without modification,
16// are permitted provided that the following conditions are met:
17//
18//   * Redistribution's of source code must retain the above copyright notice,
19//     this list of conditions and the following disclaimer.
20//
21//   * Redistribution's in binary form must reproduce the above copyright notice,
22//     this list of conditions and the following disclaimer in the documentation
23//     and/or other materials provided with the distribution.
24//
25//   * The name of Intel Corporation may not be used to endorse or promote products
26//     derived from this software without specific prior written permission.
27//
28// This software is provided by the copyright holders and contributors "as is" and
29// any express or implied warranties, including, but not limited to, the implied
30// warranties of merchantability and fitness for a particular purpose are disclaimed.
31// In no event shall the Intel Corporation or contributors be liable for any direct,
32// indirect, incidental, special, exemplary, or consequential damages
33// (including, but not limited to, procurement of substitute goods or services;
34// loss of use, data, or profits; or business interruption) however caused
35// and on any theory of liability, whether in contract, strict liability,
36// or tort (including negligence or otherwise) arising in any way out of
37// the use of this software, even if advised of the possibility of such damage.
38//
39//M*/
40
41#include "precomp.hpp"
42
43namespace cv { namespace ml {
44
45ParamGrid::ParamGrid() { minVal = maxVal = 0.; logStep = 1; }
46ParamGrid::ParamGrid(double _minVal, double _maxVal, double _logStep)
47{
48    minVal = std::min(_minVal, _maxVal);
49    maxVal = std::max(_minVal, _maxVal);
50    logStep = std::max(_logStep, 1.);
51}
52
53bool StatModel::empty() const { return !isTrained(); }
54
55int StatModel::getVarCount() const { return 0; }
56
57bool StatModel::train( const Ptr<TrainData>&, int )
58{
59    CV_Error(CV_StsNotImplemented, "");
60    return false;
61}
62
63bool StatModel::train( InputArray samples, int layout, InputArray responses )
64{
65    return train(TrainData::create(samples, layout, responses));
66}
67
68float StatModel::calcError( const Ptr<TrainData>& data, bool testerr, OutputArray _resp ) const
69{
70    Mat samples = data->getSamples();
71    int layout = data->getLayout();
72    Mat sidx = testerr ? data->getTestSampleIdx() : data->getTrainSampleIdx();
73    const int* sidx_ptr = sidx.ptr<int>();
74    int i, n = (int)sidx.total();
75    bool isclassifier = isClassifier();
76    Mat responses = data->getResponses();
77
78    if( n == 0 )
79        n = data->getNSamples();
80
81    if( n == 0 )
82        return -FLT_MAX;
83
84    Mat resp;
85    if( _resp.needed() )
86        resp.create(n, 1, CV_32F);
87
88    double err = 0;
89    for( i = 0; i < n; i++ )
90    {
91        int si = sidx_ptr ? sidx_ptr[i] : i;
92        Mat sample = layout == ROW_SAMPLE ? samples.row(si) : samples.col(si);
93        float val = predict(sample);
94        float val0 = responses.at<float>(si);
95
96        if( isclassifier )
97            err += fabs(val - val0) > FLT_EPSILON;
98        else
99            err += (val - val0)*(val - val0);
100        if( !resp.empty() )
101            resp.at<float>(i) = val;
102        /*if( i < 100 )
103        {
104            printf("%d. ref %.1f vs pred %.1f\n", i, val0, val);
105        }*/
106    }
107
108    if( _resp.needed() )
109        resp.copyTo(_resp);
110
111    return (float)(err / n * (isclassifier ? 100 : 1));
112}
113
114/* Calculates upper triangular matrix S, where A is a symmetrical matrix A=S'*S */
115static void Cholesky( const Mat& A, Mat& S )
116{
117    CV_Assert(A.type() == CV_32F);
118
119    int dim = A.rows;
120    S.create(dim, dim, CV_32F);
121
122    int i, j, k;
123
124    for( i = 0; i < dim; i++ )
125    {
126        for( j = 0; j < i; j++ )
127            S.at<float>(i,j) = 0.f;
128
129        float sum = 0.f;
130        for( k = 0; k < i; k++ )
131        {
132            float val = S.at<float>(k,i);
133            sum += val*val;
134        }
135
136        S.at<float>(i,i) = std::sqrt(std::max(A.at<float>(i,i) - sum, 0.f));
137        float ival = 1.f/S.at<float>(i, i);
138
139        for( j = i + 1; j < dim; j++ )
140        {
141            sum = 0;
142            for( k = 0; k < i; k++ )
143                sum += S.at<float>(k, i) * S.at<float>(k, j);
144
145            S.at<float>(i, j) = (A.at<float>(i, j) - sum)*ival;
146        }
147    }
148}
149
150/* Generates <sample> from multivariate normal distribution, where <mean> - is an
151   average row vector, <cov> - symmetric covariation matrix */
152void randMVNormal( InputArray _mean, InputArray _cov, int nsamples, OutputArray _samples )
153{
154    Mat mean = _mean.getMat(), cov = _cov.getMat();
155    int dim = (int)mean.total();
156
157    _samples.create(nsamples, dim, CV_32F);
158    Mat samples = _samples.getMat();
159    randu(samples, 0., 1.);
160
161    Mat utmat;
162    Cholesky(cov, utmat);
163    int flags = mean.cols == 1 ? 0 : GEMM_3_T;
164
165    for( int i = 0; i < nsamples; i++ )
166    {
167        Mat sample = samples.row(i);
168        gemm(sample, utmat, 1, mean, 1, sample, flags);
169    }
170}
171
172}}
173
174/* End of file */
175