// test_max.cpp : 定义控制台应用程序的入口点。
//
#include "stdafx.h"
#include <cv.h>
#include <highgui.h>
#include <ml.h>
#include <cxcore.h>

#include <iostream>
using namespace std;

int main()
{
 // step 1:  
//训练数据的分类标记，即4类
 float labels[16] = {1.0, 1.0,1.0,1.0,2.0,2.0,2.0,2.0,3.0,3.0,3.0,3.0,4.0,4.0,4.0,4.0};  
 CvMat labelsMat = cvMat(16, 1, CV_32FC1, labels);  
 //训练数据矩阵
 float trainingData[16][2] = { {0, 0}, {4, 1}, {4, 5}, {-1, 6},{3,11},{-2,10},{4,30},{0,25},{10,13},{15,12},{25,40},{11,35},{8,1},{9,6},{15,5},{20,-1} };  
 CvMat trainingDataMat = cvMat(16, 2, CV_32FC1, trainingData);  
 
 // step 2:  
 //训练参数设定
 CvSVMParams params;  
 params.svm_type = CvSVM::C_SVC;				 //SVM类型
 params.kernel_type = CvSVM::LINEAR;			 //核函数的类型
 
 //SVM训练过程的终止条件, max_iter:最大迭代次数  epsilon:结果的精确性
 params.term_crit = cvTermCriteria(CV_TERMCRIT_ITER, 100, FLT_EPSILON ); 
 
 // step 3:  
 //启动训练过程
 CvSVM SVM;  
 SVM.train( &trainingDataMat, &labelsMat, NULL,NULL, params); 

 // step 4:  
 //使用训练所得模型对新样本进行分类测试
 for (int i=-5; i<15; i++)  
 {  
	  for (int j=-5; j<15; j++)  
	  {  
		   float a[] = {i,j};
		   CvMat sampleMat; 
		   cvInitMatHeader(&sampleMat,1,2,CV_32FC1,a);
		   cvmSet(&sampleMat,0,0,i);										// Set M(i,j)
		   cvmSet(&sampleMat,0,1,j);										// Set M(i,j)
		   float response = SVM.predict(&sampleMat); 
		   cout<<response<<" ";
	  }
	  cout<<endl;
 } 

 // step 5:  
 //获取支持向量
 int c = SVM.get_support_vector_count();  
 cout<<endl;
 for (int i=0; i<c; i++)  
 {  
	  const float* v = SVM.get_support_vector(i);  
	  cout<<*v<<" ";
 }  
 cout<<endl;
 
 system("pause");
 return 0;
}