public function testCholesky()
{
$m1 = [[1, 0, 0, 0], [0, 1, 0, 0], [0, 0, 1, 0], [0, 0, 0, 1]];
$matrix1 = Matrix::matrixFromDoubles($m1);
$m2 = [[17, 18, 19, 20], [21, 22, 23, 24], [25, 27, 28, 29], [37, 33, 31, 30]];
$matrix2 = Matrix::matrixFromDoubles($m2);
$c = new CholeskyDecomposition($matrix1);
$c->solve($matrix2);
$mx = $c->getL();
$this->assertEquals(1.0, $mx->get(0, 0));
$this->assertEquals(1.0, $mx->get(1, 1));
$this->assertEquals(1.0, $mx->get(2, 2));
$this->assertEquals(4, $mx->getRows());
$this->assertEquals(4, $mx->getCols());
}
public function testBipolar2double()
{
// test a 1x4
$booleanData1 = [true, false, true, false];
$checkData1 = [1, -1, 1, -1];
$matrix1 = Matrix::createRowMatrix(BiPolarUtil\bipolar2double($booleanData1));
$checkMatrix1 = Matrix::createRowMatrix($checkData1);
$this->assertTrue($matrix1->equals($checkMatrix1));
// test a 2x2
$booleanData2 = [[true, false], [false, true]];
$checkData2 = [[1, -1], [-1, 1]];
$matrix2 = Matrix::matrixFromDoubles(BiPolarUtil\bipolar2double($booleanData2));
$checkMatrix2 = Matrix::matrixFromDoubles($checkData2);
$this->assertTrue($matrix2->equals($checkMatrix2));
}
/**
* Update the Hopfield weights after training.
*
* @param
* Matrix delta
* The amount to change the weights by.
*/
private function convertHopfieldMatrix(Matrix $delta)
{
// add the new weight matrix to what is there already
for ($row = 0; $row < $delta->getRows(); ++$row) {
for ($col = 0; $col < $delta->getRows(); ++$col) {
$this->addWeight($row, $col, $delta->get($row, $col));
}
}
}
public function testCopy()
{
$data = [[1.0, 2.0], [3.0, 4.0]];
$source = Matrix::matrixFromDoubles($data);
$target = new Matrix(2, 2);
MatrixMath\copy($source, $target);
$this->assertTrue($source->equals($target));
}
/**
* Randomize the matrix based on an array, modify the array.
* Previous values
* may be used, or they may be discarded, depending on the randomizer.
*
* @param
* Matrix m
* A matrix to randomize.
*/
public function randomizeMatrix(Matrix &$m)
{
$d = $m->getData();
for ($r = 0; $r < $m->getRows(); ++$r) {
for ($c = 0; $c < $m->getCols(); ++$c) {
$d[$r][$c] = $this->randomizeDouble($d[$r][$c]);
}
}
$m->setData($d);
}
/**
* Calculate the length of a vector.
*
* @param
* input
* The matrix to calculate the length of.
*
* @return Vector length.
*/
function vectorLength(Matrix $input)
{
if (!$input->isVector()) {
throw new MatrixError("Can only take the vector length of a vector.");
}
$v = $input->toPackedArray();
$rtn = 0.0;
foreach ($v as $element) {
$rtn += pow($element, 2);
}
return sqrt($rtn);
}
public function inverseCholesky()
{
$li = lowerTriangularInverse($l);
$ic = array();
for ($r = 0; $r < $this->n; ++$r) {
for ($c = 0; $c < $this->n; ++$c) {
for ($i = 0; $i < $this->n; ++$i) {
$ic[$r][$c] += $li[$i][$r] * $li[$i][$c];
}
}
}
return Matrix::matrixFromDoubles($ic);
}
public function testVectorLength()
{
$vectorData = [1.0, 2.0, 3.0, 4.0];
$vector = Matrix::createRowMatrix($vectorData);
$this->assertEquals(5, intval(MatrixMath\vectorLength($vector)));
$nonVector = new Matrix(2, 2);
try {
MatrixMath\vectorLength($nonVector);
$this->assertTrue(false);
} catch (MatrixError $e) {
}
}
