@@ -184,7 +184,7 @@ discard block |
||
184 | 184 | |
185 | 185 | /** |
186 | 186 | * @param array $records |
187 | - * @return DecisionTreeLeaf[] |
|
187 | + * @return null|DecisionTreeLeaf |
|
188 | 188 | */ |
189 | 189 | protected function getBestSplit($records) |
190 | 190 | { |
@@ -377,7 +377,7 @@ discard block |
||
377 | 377 | /** |
378 | 378 | * Used to set predefined features to consider while deciding which column to use for a split |
379 | 379 | * |
380 | - * @param array $selectedFeatures |
|
380 | + * @param integer[] $selectedFeatures |
|
381 | 381 | */ |
382 | 382 | protected function setSelectedFeatures(array $selectedFeatures) |
383 | 383 | { |
@@ -415,7 +415,6 @@ discard block |
||
415 | 415 | * each column in the given dataset. The importance values are |
416 | 416 | * normalized and their total makes 1.<br/> |
417 | 417 | * |
418 | - * @param array $labels |
|
419 | 418 | * @return array |
420 | 419 | */ |
421 | 420 | public function getFeatureImportances() |
@@ -455,7 +454,6 @@ discard block |
||
455 | 454 | * |
456 | 455 | * @param int $column |
457 | 456 | * @param DecisionTreeLeaf |
458 | - * @param array $collected |
|
459 | 457 | * |
460 | 458 | * @return array |
461 | 459 | */ |
@@ -4,9 +4,6 @@ |
||
4 | 4 | |
5 | 5 | namespace Phpml\Classification\Linear; |
6 | 6 | |
7 | -use Phpml\Helper\Predictable; |
|
8 | -use Phpml\Helper\Trainable; |
|
9 | -use Phpml\Classification\Classifier; |
|
10 | 7 | use Phpml\Classification\Linear\Perceptron; |
11 | 8 | |
12 | 9 | class Adaline extends Perceptron |
@@ -1,6 +1,6 @@ discard block |
||
1 | 1 | <?php |
2 | 2 | |
3 | -declare(strict_types=1); |
|
3 | +declare(strict_types = 1); |
|
4 | 4 | |
5 | 5 | namespace Phpml\Classification\Linear; |
6 | 6 | |
@@ -15,12 +15,12 @@ discard block |
||
15 | 15 | /** |
16 | 16 | * Batch training is the default Adaline training algorithm |
17 | 17 | */ |
18 | - const BATCH_TRAINING = 1; |
|
18 | + const BATCH_TRAINING = 1; |
|
19 | 19 | |
20 | 20 | /** |
21 | 21 | * Online training: Stochastic gradient descent learning |
22 | 22 | */ |
23 | - const ONLINE_TRAINING = 2; |
|
23 | + const ONLINE_TRAINING = 2; |
|
24 | 24 | |
25 | 25 | /** |
26 | 26 | * The function whose result will be used to calculate the network error |
@@ -52,7 +52,7 @@ discard block |
||
52 | 52 | public function __construct(float $learningRate = 0.001, int $maxIterations = 1000, |
53 | 53 | bool $normalizeInputs = true, int $trainingType = self::BATCH_TRAINING) |
54 | 54 | { |
55 | - if (! in_array($trainingType, [self::BATCH_TRAINING, self::ONLINE_TRAINING])) { |
|
55 | + if (!in_array($trainingType, [self::BATCH_TRAINING, self::ONLINE_TRAINING])) { |
|
56 | 56 | throw new \Exception("Adaline can only be trained with batch and online/stochastic gradient descent algorithm"); |
57 | 57 | } |
58 | 58 | |
@@ -104,7 +104,7 @@ discard block |
||
104 | 104 | protected function updateWeights(array $updates) |
105 | 105 | { |
106 | 106 | // Updates all weights at once |
107 | - for ($i=0; $i <= $this->featureCount; $i++) { |
|
107 | + for ($i = 0; $i <= $this->featureCount; $i++) { |
|
108 | 108 | if ($i == 0) { |
109 | 109 | $this->weights[0] += $this->learningRate * array_sum($updates); |
110 | 110 | } else { |
@@ -158,8 +158,8 @@ |
||
158 | 158 | * @param string $operator |
159 | 159 | * @param array $values |
160 | 160 | * @param array $targets |
161 | - * @param mixed $leftLabel |
|
162 | - * @param mixed $rightLabel |
|
161 | + * @param string $leftLabel |
|
162 | + * @param string $rightLabel |
|
163 | 163 | */ |
164 | 164 | protected function calculateErrorRate(float $threshold, string $operator, array $values, array $targets, $leftLabel, $rightLabel) |
165 | 165 | { |
@@ -6,9 +6,7 @@ |
||
6 | 6 | |
7 | 7 | use Phpml\Helper\Predictable; |
8 | 8 | use Phpml\Helper\Trainable; |
9 | -use Phpml\Classification\Classifier; |
|
10 | 9 | use Phpml\Classification\DecisionTree; |
11 | -use Phpml\Classification\DecisionTree\DecisionTreeLeaf; |
|
12 | 10 | |
13 | 11 | class DecisionStump extends DecisionTree |
14 | 12 | { |
@@ -1,6 +1,6 @@ discard block |
||
1 | 1 | <?php |
2 | 2 | |
3 | -declare(strict_types=1); |
|
3 | +declare(strict_types = 1); |
|
4 | 4 | |
5 | 5 | namespace Phpml\Classification\Linear; |
6 | 6 | |
@@ -122,7 +122,7 @@ discard block |
||
122 | 122 | $stepSize = ($maxValue - $minValue) / 100.0; |
123 | 123 | |
124 | 124 | $leftLabel = $this->tree->leftLeaf->classValue; |
125 | - $rightLabel= $this->tree->rightLeaf->classValue; |
|
125 | + $rightLabel = $this->tree->rightLeaf->classValue; |
|
126 | 126 | |
127 | 127 | $bestOperator = $this->tree->operator; |
128 | 128 | $bestThreshold = $this->tree->numericValue; |
@@ -130,8 +130,8 @@ discard block |
||
130 | 130 | $bestThreshold, $bestOperator, $values, $targets, $leftLabel, $rightLabel); |
131 | 131 | |
132 | 132 | foreach (['<=', '>'] as $operator) { |
133 | - for ($step = $minValue; $step <= $maxValue; $step+= $stepSize) { |
|
134 | - $threshold = (float)$step; |
|
133 | + for ($step = $minValue; $step <= $maxValue; $step += $stepSize) { |
|
134 | + $threshold = (float) $step; |
|
135 | 135 | $errorRate = $this->calculateErrorRate( |
136 | 136 | $threshold, $operator, $values, $targets, $leftLabel, $rightLabel); |
137 | 137 |
@@ -1,6 +1,6 @@ discard block |
||
1 | 1 | <?php |
2 | 2 | |
3 | -declare(strict_types=1); |
|
3 | +declare(strict_types = 1); |
|
4 | 4 | |
5 | 5 | namespace Phpml\Classification\Linear; |
6 | 6 | |
@@ -138,7 +138,7 @@ discard block |
||
138 | 138 | // Update bias |
139 | 139 | $this->weights[0] += $update * $this->learningRate; // Bias |
140 | 140 | // Update other weights |
141 | - for ($i=1; $i <= $this->featureCount; $i++) { |
|
141 | + for ($i = 1; $i <= $this->featureCount; $i++) { |
|
142 | 142 | $this->weights[$i] += $update * $sample[$i - 1] * $this->learningRate; |
143 | 143 | } |
144 | 144 | } |
@@ -190,6 +190,6 @@ discard block |
||
190 | 190 | |
191 | 191 | $predictedClass = $this->outputClass($sample); |
192 | 192 | |
193 | - return $this->labels[ $predictedClass ]; |
|
193 | + return $this->labels[$predictedClass]; |
|
194 | 194 | } |
195 | 195 | } |
@@ -1,6 +1,6 @@ discard block |
||
1 | 1 | <?php |
2 | 2 | |
3 | -declare(strict_types=1); |
|
3 | +declare(strict_types = 1); |
|
4 | 4 | |
5 | 5 | namespace Phpml\Classification\Ensemble; |
6 | 6 | |
@@ -122,7 +122,7 @@ discard block |
||
122 | 122 | // classifiers as well |
123 | 123 | $minErrorRate = 1.0; |
124 | 124 | $bestClassifier = null; |
125 | - for ($i=0; $i < $this->featureCount; $i++) { |
|
125 | + for ($i = 0; $i < $this->featureCount; $i++) { |
|
126 | 126 | $stump = new DecisionStump($i); |
127 | 127 | $stump->setSampleWeights($this->weights); |
128 | 128 | $stump->train($this->samples, $this->targets); |
@@ -185,6 +185,6 @@ discard block |
||
185 | 185 | $sum += $h * $alpha; |
186 | 186 | } |
187 | 187 | |
188 | - return $this->labels[ $sum > 0 ? 1 : -1]; |
|
188 | + return $this->labels[$sum > 0 ? 1 : -1]; |
|
189 | 189 | } |
190 | 190 | } |
@@ -1,6 +1,6 @@ discard block |
||
1 | 1 | <?php |
2 | 2 | |
3 | -declare(strict_types=1); |
|
3 | +declare(strict_types = 1); |
|
4 | 4 | |
5 | 5 | namespace Phpml\Classification\DecisionTree; |
6 | 6 | |
@@ -34,7 +34,7 @@ discard block |
||
34 | 34 | /** |
35 | 35 | * @var DecisionTreeLeaf |
36 | 36 | */ |
37 | - public $rightLeaf= null; |
|
37 | + public $rightLeaf = null; |
|
38 | 38 | |
39 | 39 | /** |
40 | 40 | * @var array |
@@ -79,7 +79,7 @@ discard block |
||
79 | 79 | |
80 | 80 | if ($this->isContinuous) { |
81 | 81 | $op = $this->operator; |
82 | - $value= $this->numericValue; |
|
82 | + $value = $this->numericValue; |
|
83 | 83 | $recordField = strval($recordField); |
84 | 84 | eval("\$result = $recordField $op $value;"); |
85 | 85 | return $result; |
@@ -100,16 +100,16 @@ discard block |
||
100 | 100 | return 0.0; |
101 | 101 | } |
102 | 102 | |
103 | - $nodeSampleCount = (float)count($this->records); |
|
103 | + $nodeSampleCount = (float) count($this->records); |
|
104 | 104 | $iT = $this->giniIndex; |
105 | 105 | |
106 | 106 | if ($this->leftLeaf) { |
107 | - $pL = count($this->leftLeaf->records)/$nodeSampleCount; |
|
107 | + $pL = count($this->leftLeaf->records) / $nodeSampleCount; |
|
108 | 108 | $iT -= $pL * $this->leftLeaf->giniIndex; |
109 | 109 | } |
110 | 110 | |
111 | 111 | if ($this->rightLeaf) { |
112 | - $pR = count($this->rightLeaf->records)/$nodeSampleCount; |
|
112 | + $pR = count($this->rightLeaf->records) / $nodeSampleCount; |
|
113 | 113 | $iT -= $pR * $this->rightLeaf->giniIndex; |
114 | 114 | } |
115 | 115 | |
@@ -133,25 +133,25 @@ discard block |
||
133 | 133 | } else { |
134 | 134 | $col = "col_$this->columnIndex"; |
135 | 135 | } |
136 | - if (! preg_match("/^[<>=]{1,2}/", $value)) { |
|
136 | + if (!preg_match("/^[<>=]{1,2}/", $value)) { |
|
137 | 137 | $value = "=$value"; |
138 | 138 | } |
139 | - $value = "<b>$col $value</b><br>Gini: ". number_format($this->giniIndex, 2); |
|
139 | + $value = "<b>$col $value</b><br>Gini: ".number_format($this->giniIndex, 2); |
|
140 | 140 | } |
141 | 141 | $str = "<table ><tr><td colspan=3 align=center style='border:1px solid;'> |
142 | 142 | $value</td></tr>"; |
143 | 143 | if ($this->leftLeaf || $this->rightLeaf) { |
144 | - $str .='<tr>'; |
|
144 | + $str .= '<tr>'; |
|
145 | 145 | if ($this->leftLeaf) { |
146 | - $str .="<td valign=top><b>| Yes</b><br>" . $this->leftLeaf->getHTML($columnNames) . "</td>"; |
|
146 | + $str .= "<td valign=top><b>| Yes</b><br>".$this->leftLeaf->getHTML($columnNames)."</td>"; |
|
147 | 147 | } else { |
148 | - $str .='<td></td>'; |
|
148 | + $str .= '<td></td>'; |
|
149 | 149 | } |
150 | - $str .='<td> </td>'; |
|
150 | + $str .= '<td> </td>'; |
|
151 | 151 | if ($this->rightLeaf) { |
152 | - $str .="<td valign=top align=right><b>No |</b><br>" . $this->rightLeaf->getHTML($columnNames) . "</td>"; |
|
152 | + $str .= "<td valign=top align=right><b>No |</b><br>".$this->rightLeaf->getHTML($columnNames)."</td>"; |
|
153 | 153 | } else { |
154 | - $str .='<td></td>'; |
|
154 | + $str .= '<td></td>'; |
|
155 | 155 | } |
156 | 156 | $str .= '</tr>'; |
157 | 157 | } |