Last active
September 18, 2022 13:09
-
-
Save unix1/7510208 to your computer and use it in GitHub Desktop.
A sample matrix inversion in PHP using Gauss-Jordan elimination. This is implementation is meant for clarity, not for performance, memory usage, or numerical stability. You can optionally turn on the debug flag to output matrix state after every iteration. See https://en.wikipedia.org/wiki/Gauss-Jordan_elimination for more info. Enjoy!
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
/** | |
* Inverts a given matrix | |
* | |
* @param array $A matrix to invert | |
* @param boolean $debug whether to print out debug info | |
* | |
* @return array inverted matrix | |
*/ | |
function invert($A, $debug = FALSE) | |
{ | |
/// @todo check rows = columns | |
$n = count($A); | |
// get and append identity matrix | |
$I = identity_matrix($n); | |
for ($i = 0; $i < $n; ++ $i) { | |
$A[$i] = array_merge($A[$i], $I[$i]); | |
} | |
if ($debug) { | |
echo "\nStarting matrix: "; | |
print_matrix($A); | |
} | |
// forward run | |
for ($j = 0; $j < $n-1; ++ $j) { | |
// for all remaining rows (diagonally) | |
for ($i = $j+1; $i < $n; ++ $i) { | |
// if the value is not already 0 | |
if ($A[$i][$j] !== 0) { | |
// adjust scale to pivot row | |
// subtract pivot row from current | |
$scalar = $A[$j][$j] / $A[$i][$j]; | |
for ($jj = $j; $jj < $n*2; ++ $jj) { | |
$A[$i][$jj] *= $scalar; | |
$A[$i][$jj] -= $A[$j][$jj]; | |
} | |
} | |
} | |
if ($debug) { | |
echo "\nForward iteration $j: "; | |
print_matrix($A); | |
} | |
} | |
// reverse run | |
for ($j = $n-1; $j > 0; -- $j) { | |
for ($i = $j-1; $i >= 0; -- $i) { | |
if ($A[$i][$j] !== 0) { | |
$scalar = $A[$j][$j] / $A[$i][$j]; | |
for ($jj = $i; $jj < $n*2; ++ $jj) { | |
$A[$i][$jj] *= $scalar; | |
$A[$i][$jj] -= $A[$j][$jj]; | |
} | |
} | |
} | |
if ($debug) { | |
echo "\nReverse iteration $j: "; | |
print_matrix($A); | |
} | |
} | |
// last run to make all diagonal 1s | |
/// @note this can be done in last iteration (i.e. reverse run) too! | |
for ($j = 0; $j < $n; ++ $j) { | |
if ($A[$j][$j] !== 1) { | |
$scalar = 1 / $A[$j][$j]; | |
for ($jj = $j; $jj < $n*2; ++ $jj) { | |
$A[$j][$jj] *= $scalar; | |
} | |
} | |
if ($debug) { | |
echo "\n1-out iteration $j: "; | |
print_matrix($A); | |
} | |
} | |
// take out the matrix inverse to return | |
$Inv = array(); | |
for ($i = 0; $i < $n; ++ $i) { | |
$Inv[$i] = array_slice($A[$i], $n); | |
} | |
return $Inv; | |
} | |
/** | |
* Prints matrix | |
* | |
* @param array $A matrix | |
* @param integer $decimals number of decimals | |
*/ | |
function print_matrix($A, $decimals = 6) | |
{ | |
foreach ($A as $row) { | |
echo "\n\t["; | |
foreach ($row as $i) { | |
echo "\t" . sprintf("%01.{$decimals}f", round($i, $decimals)); | |
} | |
echo "\t]"; | |
} | |
} | |
/** | |
* Produces an identity matrix of given size | |
* | |
* @param integer $n size of identity matrix | |
* | |
* @return array identity matrix | |
*/ | |
function identity_matrix($n) | |
{ | |
$I = array(); | |
for ($i = 0; $i < $n; ++ $i) { | |
for ($j = 0; $j < $n; ++ $j) { | |
$I[$i][$j] = ($i == $j) ? 1 : 0; | |
} | |
} | |
return $I; | |
} | |
$A = array( | |
array( 10, -15, 30, 6, -8 ), | |
array( 0, -4, 60, 11, -5 ), | |
array( 8, 9, 2, 3, 7 ), | |
array( 25, 10, -9, 9, 0 ), | |
array( 13, 3, -12, 5, 2 ), | |
); | |
echo "\nMatrix:"; | |
print_matrix($A); | |
echo "\n"; | |
$B = invert($A); | |
echo "\nInversion result:"; | |
print_matrix($B); | |
echo "\n\n"; | |
?> |
Good morning, I have a problem . I would like to inform as parameter a matrix that way , but I can not .
for ($i=0; $i <5 ; $i++) {
for ($j=0; $j< 5 ; $j++) {
$Matrix[$i][$j] = 4;}
}
$invM = invert($Matrix); <- Don't work.
Because I can only pass a parameter that way?
$Matrix = [[4,4,4,4],[4,4,4,4],[4,4,4,4],[4,4,4,4]];
$invM = invert($Matrix); <- work
Can you help me ?
Hi @rigelcastro - the matrix you specified does not have an inverse. Here's the output from octave:
octave:1> inv([4,4,4,4; 4,4,4,4; 4,4,4,4; 4,4,4,4])
warning: inverse: matrix singular to machine precision, rcond = 0
ans =
Inf Inf Inf Inf
Inf Inf Inf Inf
Inf Inf Inf Inf
Inf Inf Inf Inf
In my implementation we don't get such nice output, but instead we get a division by 0 warning from PHP. This can be made better.
Crashes on division by zero for:
$A = [
[1, 0, 7],
[0, 2, 0],
[0, 0, 3],
];
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
@pauloz1890 thanks for pointing that out! I fixed it to work with 0 values too in 2nd revision.