| 1 |
efrain |
1 |
<?php
|
|
|
2 |
|
|
|
3 |
declare(strict_types=1);
|
|
|
4 |
|
|
|
5 |
namespace Phpml\Dataset;
|
|
|
6 |
|
|
|
7 |
use Phpml\Exception\FileException;
|
|
|
8 |
|
|
|
9 |
class CsvDataset extends ArrayDataset
|
|
|
10 |
{
|
|
|
11 |
/**
|
|
|
12 |
* @var array
|
|
|
13 |
*/
|
|
|
14 |
protected $columnNames = [];
|
|
|
15 |
|
|
|
16 |
/**
|
|
|
17 |
* @throws FileException
|
|
|
18 |
*/
|
|
|
19 |
public function __construct(string $filepath, int $features, bool $headingRow = true, string $delimiter = ',', int $maxLineLength = 0)
|
|
|
20 |
{
|
|
|
21 |
if (!file_exists($filepath)) {
|
|
|
22 |
throw new FileException(sprintf('File "%s" missing.', basename($filepath)));
|
|
|
23 |
}
|
|
|
24 |
|
|
|
25 |
$handle = fopen($filepath, 'rb');
|
|
|
26 |
if ($handle === false) {
|
|
|
27 |
throw new FileException(sprintf('File "%s" can\'t be open.', basename($filepath)));
|
|
|
28 |
}
|
|
|
29 |
|
|
|
30 |
if ($headingRow) {
|
|
|
31 |
$data = fgetcsv($handle, $maxLineLength, $delimiter);
|
|
|
32 |
$this->columnNames = array_slice((array) $data, 0, $features);
|
|
|
33 |
} else {
|
|
|
34 |
$this->columnNames = range(0, $features - 1);
|
|
|
35 |
}
|
|
|
36 |
|
|
|
37 |
$samples = $targets = [];
|
|
|
38 |
while ($data = fgetcsv($handle, $maxLineLength, $delimiter)) {
|
|
|
39 |
$samples[] = array_slice($data, 0, $features);
|
|
|
40 |
$targets[] = $data[$features];
|
|
|
41 |
}
|
|
|
42 |
|
|
|
43 |
fclose($handle);
|
|
|
44 |
|
|
|
45 |
parent::__construct($samples, $targets);
|
|
|
46 |
}
|
|
|
47 |
|
|
|
48 |
public function getColumnNames(): array
|
|
|
49 |
{
|
|
|
50 |
return $this->columnNames;
|
|
|
51 |
}
|
|
|
52 |
}
|