stats_stat_correlation

(PECL stats >= 1.0.0)

stats_stat_correlation — Not documented

Description

float stats_stat_correlation ( array $arr1 , array $arr2 )

Warning

This function is currently not documented; only its argument list is available.

Parameters

arr1
arr2

Return Values

Коментарии

Mar 05

Автор: umar dot anjum at ymail dot com


I tried to use this function, but got a not-defined error. Anyway, I have created a set of functions to replace this: 



<?php



//Since Correlation needs two arrays, I am hardcoding them

$array1[0] = 59.3;

$array1[1] = 61.2;

$array1[2] = 56.8

$array1[3] = 97.55;



$array2[0] = 565.82;

$array2[1] = 54.568;

$array2[2] = 84.22;

$array2[3] = 483.55;



//To find the correlation of the two arrays, simply call the  

//function Correlation that takes two arrays:



$correlation = Correlation($array1, $array2);



//Displaying the calculated Correlation:

print $correlation;



//The functions that work behind the scene to calculate the

//correlation



function Correlation($arr1, $arr2)

{        

    $correlation = 0;

    

    $k = SumProductMeanDeviation($arr1, $arr2);

    $ssmd1 = SumSquareMeanDeviation($arr1);

    $ssmd2 = SumSquareMeanDeviation($arr2);

    

    $product = $ssmd1 * $ssmd2;

    

    $res = sqrt($product);

    

    $correlation = $k / $res;

    

    return $correlation;

}



function SumProductMeanDeviation($arr1, $arr2)

{

    $sum = 0;

    

    $num = count($arr1);

    

    for($i=0; $i<$num; $i++)

    {

        $sum = $sum + ProductMeanDeviation($arr1, $arr2, $i);

    }

    

    return $sum;

}



function ProductMeanDeviation($arr1, $arr2, $item)

{

    return (MeanDeviation($arr1, $item) * MeanDeviation($arr2, $item));

}



function SumSquareMeanDeviation($arr)

{

    $sum = 0;

    

    $num = count($arr);

    

    for($i=0; $i<$num; $i++)

    {

        $sum = $sum + SquareMeanDeviation($arr, $i);

    }

    

    return $sum;

}



function SquareMeanDeviation($arr, $item)

{

    return MeanDeviation($arr, $item) * MeanDeviation($arr, $item);

}



function SumMeanDeviation($arr)

{

    $sum = 0;

    

    $num = count($arr);

    

    for($i=0; $i<$num; $i++)

    {

        $sum = $sum + MeanDeviation($arr, $i);

    }

    

    return $sum;

}



function MeanDeviation($arr, $item)

{

    $average = Average($arr);

    

    return $arr[$item] - $average;

}    



function Average($arr)

{

    $sum = Sum($arr);

    $num = count($arr);

    

    return $sum/$num;

}



function Sum($arr)

{

    return array_sum($arr);

}



?>

2009-03-05 00:05:51

http://php5.kiev.ua/manual/ru/function.stats-stat-correlation.html

Jun 15

Автор: admin at maychu dot net (Le Cong)


Please note that this function is reserved for two arrays with continued numbers inside (just integers).

I tested this function and found that it calculate the Pearson's Correlation Coefficient of two arrays.

---

Here's suggested documentation:



stats_stat_correlation — Calculate the Pearson's Correlation Coefficient of two arrays of continued numbers.



Parameters:

arr1 = array (integer1a, interger2a ...)

arr2 = array (integer1b, interger2b ...))

(Note that the count of elements in two arrays must be equal)



Return value: Pearson's Correlation Coefficient in decimal format (ex. 0.934399822094)



Code examples:



<?php

// Provided by admin@maychu.net

$array_x = array(5,3,6,7,4,2,9,5);

$array_y = array(4,3,4,8,3,2,10,5);

$pearson = stats_stat_correlation($array_x,$array_y);

echo $pearson;

?>

2009-06-15 00:58:44

http://php5.kiev.ua/manual/ru/function.stats-stat-correlation.html

Jan 19

Автор: non at dot com


undefined for me, thus I've implemented my own correlation which is much faster and simpler than the one provided above.



function Corr($x, $y){



$length= count($x);

$mean1=array_sum($x) / $length;

$mean2=array_sum($y) / $length;



$a=0;

$b=0;

$axb=0;

$a2=0;

$b2=0;



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

{

$a=$x[$i]-$mean1;

$b=$y[$i]-$mean2;

$axb=$axb+($a*$b);

$a2=$a2+ pow($a,2);

$b2=$b2+ pow($b,2);

}



$corr= $axb / sqrt($a2*$b2);



return $corr;

}