<?php

 /*

 *KNN K-近邻方法（分类算法的实现）

 */

 /*

 *把.txt中的内容读到数组中保存,$filename:文件名称

 */

 //--------------------------------------------------------------------

 function  getFileContent($filename)

 {

     $array = array(null);

     $content = file_get_contents($filename);

     $result = explode("\r\n",$content);

     //print_r(count($result));

     for($j=0;$j<count($result);$j++)

     {

         //print_r($result[$j]."<br>");

         $con = explode(" ",$result[$j]);

         array_push($array,$con);

     }

     array_splice($array,0,1);

     return $array;

 }

 //--------------------------------------------------------------------

 /*

  */

 //希尔排序算法

 //--------------------------------------------------------------------

 function shell_sort($array)//降序

 {

     $dh=(int)(count($array)/2);

     while($dh>=1)

     {

         for($i=$dh;$i<count($array);$i++)

         {

             $temp=array($array[$i][0],$array[$i][1]);

             $j=$i-$dh;

             while($j>=0&&($array[$j][1]<$temp[1]))

             {

                 $array[$j+$dh][1]=$array[$j][1];

                 $array[$j+$dh][0]=$array[$j][0];

                 $j-=$dh;

             }

             $array[$j+$dh][1]=$temp[1];

             $array[$j+$dh][0]=$temp[0];

         }

         $dh=(int)($dh/2);

     }

     return $array;

 }

 //-------------------------------------------------------------------------

 /*

 *KNN算法

 *$test:测试文本；$train：训练文本；$flagsyes：yes；$flagsno：no

 */

 //--------------------------------------------------------------------

 function  KNN($test,$train,$flagsyes,$flagsno)

 {

     for($i=1;$i<count($train);$i++)

     {

         for($j=1;$j<count($test)-1;$j++)

         {

             if($test[$j]==$train[$i][$j]) $a[$j] = 1;

             else $a[$j] = 0;

         }

         //求两个例子之间的欧氏距离

         $sum = 0;

         for($j=1;$j<count($test)-1;$j++)

         {

             $sum += pow(1-$a[$j],2);

         }

         $distance[$i] = sqrt($sum);

     }

     $d = array(array(NULL,NULL));

     for($i=1;$i<count($train);$i++)

     {

         $d[$i-1][0] = $train[$i][count($train[$i])-1];

         $d[$i-1][1] = 1 /(1+ $distance[$i]) ;

     }

     $result =shell_sort($d);

     $k = (int) sqrt(count($train)-1);//k=sqrt(N);

     $count_yes = 0;

     for($i=0;$i<$k;$i++)

     {

         if($result[$i][0]==$flagsyes)$count_yes++;

     }

     $count_no = $k - $count_yes;

     if($count_yes>$count_no)return $flagsyes;

     else return $flagsno;

 }

 //--------------------------------------------------------------------

 $train = getFileContent("train.txt");

 $test = getFileContent("test.txt");

 for($i=1;$i<count($test);$i++)

 {

     $test[$i][count($test[0])-1] = KNN($test[$i],$train,Y,N);

 }

 /*

 *将数组中的内容读到.txt中

 */

 //--------------------------------------------------------------------

 $fp= fopen('result.txt','wb');

 for($i=0;$i<count($test);$i++)

 {

     $temp = NULL;

     for($j=0;$j<count($test[$i]);$j++)

     {

         $temp =  $test[$i][$j]."\t";

         fwrite($fp,$temp);

     }

     fwrite($fp,"\r\n");

 }

 fclose($fp);

 //--------------------------------------------------------------------

 /*

 *打印输出

 */

 //--------------------------------------------------------------------

 echo "<pre>";

 print_r($test);

 echo "</pre>";

 //--------------------------------------------------------------------

 ?>