k近邻法课后习题解答

3.2 求得的最近邻点是（2，3）

3.3 k近邻算法如下：

（思路：求得一个最近邻后，从kd树中删除这个结点）

#include <iostream>
#include <algorithm>
#include <stack>
#include <math.h>
using namespace std;
/*function of this program: build a 2d tree using the input training data
 the input is exm_set which contains a list of tuples (x,y)
 the output is a 2d tree pointer*/


struct data
{
	double x = 0;
	double y = 0;
};

struct Tnode
{
	struct data dom_elt;
    int split;
    struct Tnode * left;
    struct Tnode * right;
};

bool cmp1(data a, data b){
	return a.x < b.x;
}

bool cmp2(data a, data b){
	return a.y < b.y;
}

bool equal(data a, data b){
	if (a.x == b.x && a.y == b.y)
	{
		return true;
	}
	else{
		return false;
	}
}

void ChooseSplit(data exm_set[], int size, int &split, data &SplitChoice){
	/*compute the variance on every dimension. Set split as the dismension that have the biggest
     variance. Then choose the instance which is the median on this split dimension.*/
	/*compute variance on the x,y dimension. DX=EX^2-(EX)^2*/
    double tmp1,tmp2;
    tmp1 = tmp2 = 0;
    for (int i = 0; i < size; ++i)
    {
    	tmp1 += 1.0 / (double)size * exm_set[i].x * exm_set[i].x;
    	tmp2 += 1.0 / (double)size * exm_set[i].x;
    }
    double v1 = tmp1 - tmp2 * tmp2;  //compute variance on the x dimension
    
    tmp1 = tmp2 = 0;
    for (int i = 0; i < size; ++i)
    {
    	tmp1 += 1.0 / (double)size * exm_set[i].y * exm_set[i].y;
    	tmp2 += 1.0 / (double)size * exm_set[i].y;
    }
    double v2 = tmp1 - tmp2 * tmp2;  //compute variance