PowerStationDatasetSplitter.php

<?php

namespace App\Application\ML;

use App\Application\Path\AppPathResolver;
use App\Domain\FileNames;
use League\Csv\CannotInsertRecord;
use League\Csv\Exception;
use League\Csv\UnavailableStream;
use League\Csv\Writer;
use Rubix\ML\Datasets\Unlabeled;
use Rubix\ML\Extractors\CSV;
use Rubix\ML\Transformers\NumericStringConverter;

readonly class PowerStationDatasetSplitter
{
    public function __construct(
        private AppPathResolver $appPathResolver,
    ) {
    }

    /**
     * @throws UnavailableStream
     * @throws CannotInsertRecord
     * @throws Exception
     */
    public function split(): void
    {
        $dataset = Unlabeled::fromIterator(new CSV(
            $this->appPathResolver->getResourcesPath(FileNames::VALID_DATA),
            header: true,
        ))->apply(new NumericStringConverter());

        [$training, $testing] = $dataset->randomize()->split(0.8);

        $writer = Writer::createFromPath($this->appPathResolver->getResourcesPath(FileNames::TRAINING_SET), 'w');
        foreach ($training as $train) {
            $writer->insertOne($train);
        }

        $writer = Writer::createFromPath($this->appPathResolver->getResourcesPath(FileNames::TEST_SET), 'w');
        foreach ($testing as $test) {
            $writer->insertOne($test);
        }
    }
}