public function testAccuracy()
{
// Arrange
$actual = [true, true, false, false, true];
$predicted = [true, false, true, false, false];
$testSet = [null, null, null, null, null];
$results = new RunnerResults($predicted, $actual, $testSet);
// Act
$accuracy = $results->getAccuracy();
// Assert
$this->assertEquals(2.0 / 5.0, $accuracy);
}
public function run(array $dataSet, array $dataSetLabels)
{
// Generate folds
$partitioner = new CrossValidationRandomPartitioner(new DataSet($dataSet, $dataSetLabels), $this->numberOfPartitions);
$resultsList = [];
for ($i = 0; $i < $this->numberOfPartitions; $i++) {
// For each fold, train and classify
$currentPartition = $partitioner->getPartition($i);
$currentTest = $currentPartition->getTestData();
$currentTraining = $currentPartition->getTrainingData();
$classifier = new $this->classifierClassName($currentTraining->getData(), $currentTraining->getLabels());
$predicted = $this->classify($classifier, $currentTest->getData());
$resultsList[] = new RunnerResults($predicted, $currentTest->getLabels(), $currentTest->getData());
}
return RunnerResults::combine($resultsList);
}
public function format(RunnerResults $results)
{
return sprintf(self::$format, count($results->getTestSet()), $results->getFalsePositives(), $results->getFalseNegatives(), $results->getPrecision(), $results->getRecall(), $results->getSpecificity(), $results->getAccuracy(), $results->getFMeasure());
}
