(PHP 5 >= 5.1.0)
array_product — Ermittelt das Produkt von Werten in einem Array
array_product() gibt das Produkt von Werten in einem Array.
Das Array.
Gibt das Produkt als Integer oder Float zurück.
Beispiel #1 array_product() Beispiel
<?php
$a = array(2, 4, 6, 8);
echo "Produkt(a) = " . array_product($a) . "\n";
?>
Das oben gezeigte Beispiel erzeugt folgende Ausgabe:
Produkt(a) = 384
You can use array_product to calculate the factorial of n:
<?php
function factorial( $n )
{
if( $n < 1 ) $n = 1;
return array_product( range( 1, $n ));
}
?>
If you need the factorial without having array_product available, here is one:
<?php
function factorial( $n )
{
if( $n < 1 ) $n = 1;
for( $p++; $n; ) $p *= $n--;
return $p;
}
?>
gmail @ algofoogle is right, so we can extend our own array_product() to flexibly accept the empty product value. Zero (0) is the default (to be compatible with PHP behavior), but you could change this to 1 for mathematical purposes or null for logical.
<?php
if (! function_exists('array_product')) {
function array_product($array, $emptyProduct = 0) {
if (is_array($array)) {
return (0 == count($array) ? $emptyProduct : array_reduce($array, '_array_product', 1));
} else {
trigger_error('Param #1 must be an array', E_USER_ERROR);
return false;
}
}
function _array_product($v,$w) { return $v * $w; }
}
?>
Here is how you can multiply two arrays in the form of matrixes using a bit of matrix algebra (M*M).
By calling the function multiplyMatrix, you will be multiplying two sparse matrixes (zeros need not be included in the array for the operation to be performed).
<?php
$M = array(
0=>array(1=>1,4=>1),
1=>array(2=>1,3=>1),
3=>array(1=>1),
4=>array(5=>1),
5=>array(6=>1)
);
$M1 = multiplyMatrix($M, $M); //multiplying $M by itself
echo '<pre>';print_r($M1);echo '</pre>';
function multiplyMatrix($M1, $M2)
{
#Helena F Deus, Oct 06, 2008
##Multiply two matrixes; $M1 and $M2 can be sparse matrixes, the indexes on both should match
if(is_file($M1)) {$matrix1 = unserialize(file_get_contents($M1));}
else $matrix1 = $M1;
#transpose M2
$M2t = transpose($M2);
foreach ($M2t as $row=>$tmp) {
##sum the result of the value in the col multiplied by the value in the vector on the corresponding row
foreach ($M1 as $row1=>$tmp1) {
$multiply[$row1] = array_rproduct($tmp,$tmp1);
if(!$multiply[$row1]){
exit;
}
}
foreach ($multiply as $row1=>$vals) {
$sum[$row][$row1]=array_sum($vals);
}
}
$r=transpose($sum);
return ($r);
}
function transpose($M)
{
foreach ($M as $row=>$cols) {
foreach ($cols as $col=>$value) {
if($value)
$Mt[$col][$row]=$value;
}
}
ksort($Mt);
return ($Mt);
}
function array_rproduct($a1, $a2)
{
foreach ($a1 as $line=>$cols) {
$a3[$line] = $a1[$line]*$a2[$line];
foreach ($a2 as $line2=>$cols2) {
$a3[$line2] = $a1[$line2]*$a2[$line2];
}
}
ksort($a3);
return ($a3);
}
?>
Just in relation to "bishop" and the overall behaviour of array_product... The "empty product" (i.e. product of no values) is supposed to be defined as "1":
http://en.wikipedia.org/wiki/Empty_product
...however PHP's array_product() returns int(0) if it is given an empty array. bishop's code does this, too (so it IS a compatible replacement). Ideally, array_product() should probably return int(1). I guess it depends on your specific context or rationale.
You might normally presume int(0) to be a suitable return value if there are no inputs, but let's say that you're calculating a price based on "percentage" offsets:
$price = 10.0;
$discounts = get_array_of_customer_discounts();
$price = $price * array_product($discounts);
...if there are NO "discounts", the price will come out as 0, instead of 10.0
An observation about the _use_ of array_product with primes:
$a=$arrayOfSomePrimes=(2,3,11);
// 2 being the first prime (these days)
$codeNum=array_product($a); // gives 66 (== 2*3*11)
echo "unique product(\$a) = " . array_product($a) . "\n";
The 66 can (only) be split into its original primes,
which can be transformed into their place in the row of primes (2,3,5,7,11,13,17,19...) giving (1,2,3,4,5,6,7,8...)
The 66 gives the places {1,2,5} in the row of primes. The number "66" is unique as a code for {1,2,5}
So you can define the combination of table-columns {1,2,5} in "66". The bigger the combination, the more efficient in memory/transmission, the less in calculation.
Yet another implementation of array_product() using PHP's native array_reduce():
if (! function_exists('array_product')) {
function array_product($array) {
if (is_array($array)) {
return (0 == count($array) ? 0 : array_reduce($array, '_array_product', 1));
} else {
trigger_error('Param #1 must be an array', E_USER_ERROR);
return false;
}
}
function _array_product($v,$w) { return $v * $w; }
}
Regarding Andre D function to test if all values in an array of booleans are true, you can also use:
<?php
$allTrue = (! in_array(false, $arrayToCheck));
?>
Both this method and Andre D's are O(n), but this method has a lower k in the average case: in_array() stops once it finds the first false, while array_product must always traverse the entire array.
if you don't have PHP 5.xx . you can use this function.
It does not make sure that the variables are numeric.
function calculate_array_product($array="")
{
if(is_array($array))
{
foreach($array as $key => $value)
{
$productkey = $productkey + $key;
}
return $productkey;
}
return NULL;
}
This function can be used to test if all values in an array of booleans are TRUE.
Consider:
<?php
function outbool($test)
{
return (bool) $test;
}
$check[] = outbool(TRUE);
$check[] = outbool(1);
$check[] = outbool(FALSE);
$check[] = outbool(0);
$result = (bool) array_product($check);
// $result is set to FALSE because only two of the four values evaluated to TRUE
?>
The above is equivalent to:
<?php
$check1 = outbool(TRUE);
$check2 = outbool(1);
$check3 = outbool(FALSE);
$check4 = outbool(0);
$result = ($check1 && $check2 && $check3 && $check4);
?>
This use of array_product is especially useful when testing an indefinite number of booleans and is easy to construct in a loop.
If you don't have PHP 5
$ar = array(1,2,3,4);
$t = 1;
foreach($ar as $n){
$t *= $n;
}
echo $t; //output: 24