Transform labels now returns self for method chaining

Author: andrewdalpino

CHANGELOG.md

@@ -10,6 +10,7 @@
     - Removed Robust Z Score tolerance parameter
     - Added slice method to Dataset API
     - Loda now performs density estimation on the fly
+    - Transform labels now returns self for method chaining
 
 - 0.0.12-beta
     - Added AdaMax neural network Optimizer

docs/datasets/labeled.md

@@ -106,9 +106,9 @@ array(2) {
 ```
 
 #### Transform
-Transform the labels in the dataset using a callback function:
+Transform the labels in the dataset using a callback function and return self for method chaining:
 ```php
-public transformLabels(callable $fn) : void
+public transformLabels(callable $fn) : self
 ```
 
 > **Note:** The callback function is given a label as its only argument and should return the transformed label as a continuous or categorical value.
@@ -158,7 +158,7 @@ $filtered = $dataset->filterByLabel(function ($label)) {
 ```
 
 #### Sorting
-Sort the dataset by label:
+Sort the dataset by label and return self for method chaining:
 ```php
 public sortByLabel(bool $descending = false) : self
 ```

src/AnomalyDetectors/KDLOF.php

@@ -216,9 +216,8 @@ public function predict(Dataset $dataset) : array
      */
     public function rank(Dataset $dataset) : array
     {
-        if ($this->tree->bare()) {
-            throw new RuntimeException('The estimator has not'
-                . ' been trained.');
+        if ($this->tree->bare() or empty($this->lrds)) {
+            throw new RuntimeException('The estimator has not been trained.');
         }
 
         DatasetIsCompatibleWithEstimator::check($dataset, $this);
@@ -235,11 +234,6 @@ public function rank(Dataset $dataset) : array
      */
     protected function localOutlierFactor(array $sample) : float
     {
-        if (empty($this->lrds)) {
-            throw new RuntimeException('Local reachability distances have'
-                . ' not been computed, must train estimator first.');
-        }
-
         [$indices, $distances] = $this->tree->nearest($sample, $this->k);
 
         $lrd = $this->localReachabilityDensity($indices, $distances);

src/Datasets/Labeled.php

@@ -192,12 +192,13 @@ public function labelType() : ?int
     }
 
     /**
-     * Map labels to their new values.
+     * Map labels to their new valuesa dn return self for method chaining.
      *
      * @param callable $callback
      * @throws \RuntimeException
+     * @return self
      */
-    public function transformLabels(callable $callback) : void
+    public function transformLabels(callable $callback) : self
     {
         $labels = array_map($callback, $this->labels);
 
@@ -209,6 +210,8 @@ public function transformLabels(callable $callback) : void
         }
 
         $this->labels = $labels;
+
+        return $this;
     }
 
     /**
@@ -445,7 +448,7 @@ public function filterByLabel(callable $callback) : self
      * @param bool $descending
      * @return self
      */
-    public function sortByColumn(int $index, bool $descending = false)
+    public function sortByColumn(int $index, bool $descending = false) : self
     {
         $order = $this->column($index);
 
@@ -463,9 +466,9 @@ public function sortByColumn(int $index, bool $descending = false)
      * Sort the dataset in place by its labels.
      *
      * @param bool $descending
-     * @return \Rubix\ML\Datasets\Dataset
+     * @return self
      */
-    public function sortByLabel(bool $descending = false) : Dataset
+    public function sortByLabel(bool $descending = false) : self
     {
         array_multisort(
             $this->labels,

src/Datasets/Unlabeled.php

@@ -257,7 +257,7 @@ public function filterByColumn(int $index, callable $callback) : self
      * @param bool $descending
      * @return self
      */
-    public function sortByColumn(int $index, bool $descending = false)
+    public function sortByColumn(int $index, bool $descending = false) : self
     {
         $column = $this->column($index);