php-ml/tests/Preprocessing/ImputerTest.php

177 lines
4.4 KiB
PHP
Raw Normal View History

<?php
2016-11-20 21:53:17 +00:00
declare(strict_types=1);
namespace Phpml\Tests\Preprocessing;
use Phpml\Preprocessing\Imputer;
use Phpml\Preprocessing\Imputer\Strategy\MeanStrategy;
use Phpml\Preprocessing\Imputer\Strategy\MedianStrategy;
use Phpml\Preprocessing\Imputer\Strategy\MostFrequentStrategy;
2017-02-03 11:58:25 +00:00
use PHPUnit\Framework\TestCase;
2017-02-03 11:58:25 +00:00
class ImputerTest extends TestCase
{
public function testComplementsMissingValuesWithMeanStrategyOnColumnAxis(): void
{
$data = [
[1, null, 3, 4],
[4, 3, 2, 1],
[null, 6, 7, 8],
[8, 7, null, 5],
];
$imputeData = [
[1, 5.33, 3, 4],
[4, 3, 2, 1],
[4.33, 6, 7, 8],
[8, 7, 4, 5],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MeanStrategy(), Imputer::AXIS_COLUMN, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data, '', $delta = 0.01);
}
public function testComplementsMissingValuesWithMeanStrategyOnRowAxis(): void
{
$data = [
[1, null, 3, 4],
[4, 3, 2, 1],
[null, 6, 7, 8],
[8, 7, null, 5],
];
$imputeData = [
[1, 2.66, 3, 4],
[4, 3, 2, 1],
[7, 6, 7, 8],
[8, 7, 6.66, 5],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MeanStrategy(), Imputer::AXIS_ROW, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data, '', $delta = 0.01);
}
public function testComplementsMissingValuesWithMediaStrategyOnColumnAxis(): void
{
$data = [
[1, null, 3, 4],
[4, 3, 2, 1],
[null, 6, 7, 8],
[8, 7, null, 5],
];
$imputeData = [
[1, 6, 3, 4],
[4, 3, 2, 1],
[4, 6, 7, 8],
[8, 7, 3, 5],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MedianStrategy(), Imputer::AXIS_COLUMN, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data, '', $delta = 0.01);
}
public function testComplementsMissingValuesWithMediaStrategyOnRowAxis(): void
{
$data = [
[1, null, 3, 4],
[4, 3, 2, 1],
[null, 6, 7, 8],
[8, 7, null, 5],
];
$imputeData = [
[1, 3, 3, 4],
[4, 3, 2, 1],
[7, 6, 7, 8],
[8, 7, 7, 5],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MedianStrategy(), Imputer::AXIS_ROW, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data, '', $delta = 0.01);
}
public function testComplementsMissingValuesWithMostFrequentStrategyOnColumnAxis(): void
{
$data = [
[1, null, 3, 4],
[4, 3, 2, 1],
[null, 6, 7, 8],
[8, 7, null, 5],
[8, 3, 2, 5],
];
$imputeData = [
[1, 3, 3, 4],
[4, 3, 2, 1],
[8, 6, 7, 8],
[8, 7, 2, 5],
[8, 3, 2, 5],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MostFrequentStrategy(), Imputer::AXIS_COLUMN, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data);
}
public function testComplementsMissingValuesWithMostFrequentStrategyOnRowAxis(): void
{
$data = [
[1, null, 3, 4, 3],
[4, 3, 2, 1, 7],
[null, 6, 7, 8, 6],
[8, 7, null, 5, 5],
[8, 3, 2, 5, 4],
];
$imputeData = [
[1, 3, 3, 4, 3],
[4, 3, 2, 1, 7],
[6, 6, 7, 8, 6],
[8, 7, 5, 5, 5],
[8, 3, 2, 5, 4],
];
2016-06-16 22:16:49 +00:00
$imputer = new Imputer(null, new MostFrequentStrategy(), Imputer::AXIS_ROW, $data);
$imputer->transform($data);
$this->assertEquals($imputeData, $data);
}
public function testImputerWorksOnFitSamples(): void
{
$trainData = [
[1, 3, 4],
[6, 7, 8],
[8, 7, 5],
];
$data = [
[1, 3, null],
[6, null, 8],
[null, 7, 5],
];
$imputeData = [
[1, 3, 5.66],
[6, 5.66, 8],
[5, 7, 5],
];
$imputer = new Imputer(null, new MeanStrategy(), Imputer::AXIS_COLUMN, $trainData);
$imputer->transform($data);
$this->assertEquals($imputeData, $data, '', $delta = 0.01);
}
}