396 convert regressors to num power

.github/workflows/ci.yml

@@ -81,6 +81,7 @@ jobs:
         run: composer analyze-ci
 
       - name: Unit Tests
+        #run: vendor/bin/phpunit --display-warning --display-deprecations --display-notices --testsuite="Anomaly Detectors,Backends,Base,Classifiers,Clusterers,Cross Validation,Datasets,Extractors,Graph,Helpers,Kernels,Loggers,NeuralNet,Persisters,Regressors,Serializers,Specifications,Strategies,Tokenizers,Transformers"
         run: composer test
 
       - name: Check Coding Style

CHANGELOG.md

@@ -9,6 +9,10 @@
     - RBX Serializer only tracks major library version number
     - Convert NeuralNet classes to use NDArray instead of Matrix
     - Converted Network back from a class to an interface
+    - Added array_pack() function to replace array_map('array_values', $samples)
+    - Converted Regressor classes to use NDArray instead of Matrix
+    - Added benchmark tests for Activation Functions, based on NDArray
+    - Added benchmark tests for Regressors, based on NDArray
 
 - 2.5.0
     - Added Vantage Point Spatial tree

benchmarks/NeuralNet/ActivationFunctions/ELU/ELUBench.php

@@ -0,0 +1,58 @@
+<?php
+
+namespace Rubix\ML\Benchmarks\NeuralNet\ActivationFunctions\ELU;
+
+use NDArray;
+use NumPower;
+use Rubix\ML\NeuralNet\ActivationFunctions\ELU\ELU;
+
+/**
+ * @Groups({"ActivationFunctions"})
+ * @BeforeMethods({"setUp"})
+ */
+class ELUBench
+{
+    /**
+     * @var NDArray
+     */
+    protected NDArray $z;
+
+    /**
+     * @var NDArray
+     */
+    protected NDArray $computed;
+
+    /**
+     * @var ELU
+     */
+    protected ELU $activationFn;
+
+    public function setUp() : void
+    {
+        $this->z = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->computed = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->activationFn = new ELU();
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function activate() : void
+    {
+        $this->activationFn->activate($this->z);
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function differentiate() : void
+    {
+        $this->activationFn->differentiate($this->z, $this->computed);
+    }
+}

benchmarks/NeuralNet/ActivationFunctions/ELUBench.php

@@ -14,17 +14,17 @@ class ELUBench
     /**
      * @var Matrix
      */
-    protected $z;
+    protected Matrix $z;
 
     /**
      * @var Matrix
      */
-    protected $computed;
+    protected Matrix $computed;
 
     /**
      * @var ELU
      */
-    protected $activationFn;
+    protected ELU $activationFn;
 
     public function setUp() : void
     {

benchmarks/NeuralNet/ActivationFunctions/GELU/GELUBench.php

@@ -0,0 +1,58 @@
+<?php
+
+namespace Rubix\ML\Benchmarks\NeuralNet\ActivationFunctions\GELU;
+
+use NDArray;
+use NumPower;
+use Rubix\ML\NeuralNet\ActivationFunctions\GELU\GELU;
+
+/**
+ * @Groups({"ActivationFunctions"})
+ * @BeforeMethods({"setUp"})
+ */
+class GELUBench
+{
+    /**
+     * @var NDArray
+     */
+    protected NDArray $z;
+
+    /**
+     * @var NDArray
+     */
+    protected NDArray $computed;
+
+    /**
+     * @var GELU
+     */
+    protected GELU $activationFn;
+
+    public function setUp() : void
+    {
+        $this->z = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->computed = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->activationFn = new GELU();
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function activate() : void
+    {
+        $this->activationFn->activate($this->z);
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function differentiate() : void
+    {
+        $this->activationFn->differentiate($this->z);
+    }
+}

benchmarks/NeuralNet/ActivationFunctions/GELUBench.php

@@ -40,7 +40,7 @@ public function setUp() : void
      * @Iterations(3)
      * @OutputTimeUnit("milliseconds", precision=3)
      */
-    public function compute() : void
+    public function activate() : void
     {
         $this->activationFn->activate($this->z);
     }

benchmarks/NeuralNet/ActivationFunctions/HyperbolicTangent/HyperbolicTangentBench.php

@@ -0,0 +1,58 @@
+<?php
+
+namespace Rubix\ML\Benchmarks\NeuralNet\ActivationFunctions\HyperbolicTangent;
+
+use NDArray;
+use NumPower;
+use Rubix\ML\NeuralNet\ActivationFunctions\HyperbolicTangent\HyperbolicTangent;
+
+/**
+ * @Groups({"ActivationFunctions"})
+ * @BeforeMethods({"setUp"})
+ */
+class HyperbolicTangentBench
+{
+    /**
+     * @var NDArray
+     */
+    protected NDArray $z;
+
+    /**
+     * @var NDArray
+     */
+    protected NDArray $computed;
+
+    /**
+     * @var HyperbolicTangent
+     */
+    protected HyperbolicTangent $activationFn;
+
+    public function setUp() : void
+    {
+        $this->z = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->computed = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->activationFn = new HyperbolicTangent();
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function activate() : void
+    {
+        $this->activationFn->activate($this->z);
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function differentiate() : void
+    {
+        $this->activationFn->differentiate($this->computed);
+    }
+}

benchmarks/NeuralNet/ActivationFunctions/LeakyReLU/LeakyReLUBench.php

@@ -0,0 +1,58 @@
+<?php
+
+namespace Rubix\ML\Benchmarks\NeuralNet\ActivationFunctions\LeakyReLU;
+
+use NDArray;
+use NumPower;
+use Rubix\ML\NeuralNet\ActivationFunctions\LeakyReLU\LeakyReLU;
+
+/**
+ * @Groups({"ActivationFunctions"})
+ * @BeforeMethods({"setUp"})
+ */
+class LeakyReLUBench
+{
+    /**
+     * @var NDArray
+     */
+    protected NDArray $z;
+
+    /**
+     * @var NDArray
+     */
+    protected NDArray $computed;
+
+    /**
+     * @var LeakyReLU
+     */
+    protected LeakyReLU $activationFn;
+
+    public function setUp() : void
+    {
+        $this->z = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->computed = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->activationFn = new LeakyReLU();
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function activate() : void
+    {
+        $this->activationFn->activate($this->z);
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function differentiate() : void
+    {
+        $this->activationFn->differentiate($this->z);
+    }
+}

benchmarks/NeuralNet/ActivationFunctions/ReLU/ReLUBench.php

@@ -0,0 +1,58 @@
+<?php
+
+namespace Rubix\ML\Benchmarks\NeuralNet\ActivationFunctions\ReLU;
+
+use NDArray;
+use NumPower;
+use Rubix\ML\NeuralNet\ActivationFunctions\ReLU\ReLU;
+
+/**
+ * @Groups({"ActivationFunctions"})
+ * @BeforeMethods({"setUp"})
+ */
+class ReLUBench
+{
+    /**
+     * @var NDArray
+     */
+    protected NDArray $z;
+
+    /**
+     * @var NDArray
+     */
+    protected NDArray $computed;
+
+    /**
+     * @var ReLU
+     */
+    protected ReLU $activationFn;
+
+    public function setUp() : void
+    {
+        $this->z = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->computed = NumPower::uniform([500, 500], low: -1.0, high: 1.0);
+
+        $this->activationFn = new ReLU();
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function activate() : void
+    {
+        $this->activationFn->activate($this->z);
+    }
+
+    /**
+     * @Subject
+     * @Iterations(3)
+     * @OutputTimeUnit("milliseconds", precision=3)
+     */
+    public function differentiate() : void
+    {
+        $this->activationFn->differentiate($this->z);
+    }
+}