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

composer.json

@@ -38,6 +38,7 @@
         "andrewdalpino/okbloomer": "^1.0",
         "psr/log": "^1.1|^2.0|^3.0",
         "rubix/tensor": "^3.0",
+        "rubixml/numpower": "dev-main",
         "symfony/polyfill-mbstring": "^1.0",
         "symfony/polyfill-php80": "^1.17",
         "symfony/polyfill-php82": "^1.27",
@@ -52,7 +53,8 @@
         "phpstan/phpstan": "^2.0",
         "phpstan/phpstan-phpunit": "^2.0",
         "phpunit/phpunit": "^12.0",
-        "swoole/ide-helper": "^5.1"
+        "swoole/ide-helper": "^5.1",
+        "apphp/pretty-print": "^0.6.0"
     },
     "suggest": {
         "ext-tensor": "For fast Matrix/Vector computing",

docs/datasets/generators/hyperplane.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Datasets/Generators/Hyperplane.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Datasets/Generators/Hyperplane/Hyperplane.php">[source]</a></span>
 
 # Hyperplane
 Generates a labeled dataset whose samples form a hyperplane in n-dimensional vector space and whose labels are continuous values drawn from a uniform random distribution between -1 and 1. When the number of coefficients is either 1, 2 or 3, the samples form points, lines, and planes respectively. Due to its linearity, Hyperplane is especially useful for testing linear regression models.
@@ -16,7 +16,7 @@ Generates a labeled dataset whose samples form a hyperplane in n-dimensional vec
 
 ## Example
 ```php
-use Rubix\ML\Datasets\Generators\Hyperplane;
+use Rubix\ML\Datasets\Generators\Hyperplane\Hyperplane;
 
 $generator = new Hyperplane([0.1, 3, -5, 0.01], 150.0, 0.25);
 ```

docs/datasets/generators/swiss-roll.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Datasets/Generators/SwissRoll.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Datasets/Generators/SwissRoll/SwissRoll.php">[source]</a></span>
 
 # Swiss Roll
 Generate a non-linear 3-dimensional dataset resembling a *swiss roll* or spiral. The labels are the seeds to the swiss roll transformation.
@@ -19,7 +19,7 @@ Generate a non-linear 3-dimensional dataset resembling a *swiss roll* or spiral.
 
 ## Example
 ```php
-use Rubix\ML\Datasets\Generators\SwissRoll;
+use Rubix\ML\Datasets\Generators\SwissRoll\SwissRoll;
 
 $generator = new SwissRoll(5.5, 1.5, -2.0, 10, 21.0, 0.2);
 ```

docs/regressors/adaline.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/Adaline.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/Adaline/Adaline.php">[source]</a></span>
 
 # Adaline
 *Adaptive Linear Neuron* is a single layer feed-forward neural network with a continuous linear output neuron suitable for regression tasks. Training is equivalent to solving L2 regularized linear regression ([Ridge](ridge.md)) online using Mini Batch Gradient Descent.
@@ -20,9 +20,9 @@
 
 ## Example
 ```php
-use Rubix\ML\Regressors\Adaline;
-use Rubix\ML\NeuralNet\Optimizers\Adam;
-use Rubix\ML\NeuralNet\CostFunctions\HuberLoss;
+use Rubix\ML\Regressors\Adaline\Adaline;
+use Rubix\ML\NeuralNet\Optimizers\Adam\Adam;
+use Rubix\ML\NeuralNet\CostFunctions\HuberLoss\HuberLoss;
 
 $estimator = new Adaline(256, new Adam(0.001), 1e-4, 500, 1e-6, 5, new HuberLoss(2.5));
 ```

docs/regressors/extra-tree-regressor.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/ExtraTreeRegressor.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/ExtraTreeRegressor/ExtraTreeRegressor.php">[source]</a></span>
 
 # Extra Tree Regressor
 *Extremely Randomized* Regression Trees differ from standard [Regression Trees](regression-tree.md) in that they choose candidate splits at random rather than searching the entire feature column for the best value to split on. Extra Trees are also faster to build and their predictions have higher variance than a regular decision tree regressor.
@@ -17,7 +17,7 @@
 
 ## Example
 ```php
-use Rubix\ML\Regressors\ExtraTreeRegressor;
+use Rubix\ML\Regressors\ExtraTreeRegressor\ExtraTreeRegressor;
 
 $estimator = new ExtraTreeRegressor(30, 5, 0.05, null);
 ```

docs/regressors/gradient-boost.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/GradientBoost.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/GradientBoost/GradientBoost.php">[source]</a></span>
 
 # Gradient Boost
 Gradient Boost (GBM) is a stage-wise additive ensemble that uses a Gradient Descent boosting scheme for training boosters (Decision Trees) to correct the error residuals of a base learner.
@@ -28,8 +28,8 @@ Gradient Boost (GBM) is a stage-wise additive ensemble that uses a Gradient Desc
 
 ## Example
 ```php
-use Rubix\ML\Regressors\GradientBoost;
-use Rubix\ML\Regressors\RegressionTree;
+use Rubix\ML\Regressors\GradientBoost\GradientBoost;
+use Rubix\ML\Regressors\RegressionTree\RegressionTree;
 use Rubix\ML\CrossValidation\Metrics\SMAPE;
 
 $estimator = new GradientBoost(new RegressionTree(3), 0.1, 0.8, 1000, 1e-4, 3, 10, 0.1, new SMAPE());

docs/regressors/kd-neighbors-regressor.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/KDNeighborsRegressor.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/KDNeighborsRegressor/KDNeighborsRegressor.php">[source]</a></span>
 
 # K-d Neighbors Regressor
 A fast implementation of [KNN Regressor](knn-regressor.md) using a spatially-aware binary tree for nearest neighbors search. K-d Neighbors Regressor works by locating the neighborhood of a sample via binary search and then does a brute force search only on the samples close to or within the neighborhood of the unknown sample. The main advantage of K-d Neighbors over brute force KNN is inference speed, however, it cannot be partially trained.
@@ -16,7 +16,7 @@ A fast implementation of [KNN Regressor](knn-regressor.md) using a spatially-awa
 
 ## Example
 ```php
-use Rubix\ML\Regressors\KDNeighborsRegressor;
+use Rubix\ML\Regressors\KDNeighborsRegressor\KDNeighborsRegressor;
 use Rubix\ML\Graph\Trees\BallTree;
 
 $estimator = new KDNeighborsRegressor(20, true, new BallTree(50));

docs/regressors/knn-regressor.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/KNNRegressor.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/KNNRegressor/KNNRegressor.php">[source]</a></span>
 
 # KNN Regressor
 K Nearest Neighbors (KNN) is a brute-force distance-based learner that locates the k nearest training samples from the training set and averages their labels to make a prediction. K Nearest Neighbors (KNN) is considered a *lazy* learner because it performs most of its computation at inference time.
@@ -19,7 +19,7 @@ K Nearest Neighbors (KNN) is a brute-force distance-based learner that locates t
 
 ## Example
 ```php
-use Rubix\ML\Regressors\KNNRegressor;
+use Rubix\ML\Regressors\KNNRegressor\KNNRegressor;
 use Rubix\ML\Kernels\Distance\SafeEuclidean;
 
 $estimator = new KNNRegressor(5, false, new SafeEuclidean());

docs/regressors/mlp-regressor.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/MLPRegressor.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/MLPRegressor/MLPRegressor.php">[source]</a></span>
 
 # MLP Regressor
 A multilayer feed-forward neural network with a continuous output layer suitable for regression problems. The Multilayer Perceptron regressor is able to handle complex non-linear regression problems by forming higher-order representations of the input features using intermediate user-defined hidden layers. The MLP also has network snapshotting and progress monitoring to ensure that the model achieves the highest validation score per a given training time budget.
@@ -26,12 +26,12 @@ A multilayer feed-forward neural network with a continuous output layer suitable
 
 ## Example
 ```php
-use Rubix\ML\Regressors\MLPRegressor;
-use Rubix\ML\NeuralNet\CostFunctions\LeastSquares;
-use Rubix\ML\NeuralNet\Layers\Dense;
-use Rubix\ML\NeuralNet\Layers\Activation;
-use Rubix\ML\NeuralNet\ActivationFunctions\ReLU;
-use Rubix\ML\NeuralNet\Optimizers\RMSProp;
+use Rubix\ML\Regressors\MLPRegressor\MLPRegressor;
+use Rubix\ML\NeuralNet\CostFunctions\LeastSquares\LeastSquares;
+use Rubix\ML\NeuralNet\Layers\Dense\Dense;
+use Rubix\ML\NeuralNet\Layers\Activation\Activation;
+use Rubix\ML\NeuralNet\ActivationFunctions\ReLU\ReLU;
+use Rubix\ML\NeuralNet\Optimizers\RMSProp\RMSProp;
 use Rubix\ML\CrossValidation\Metrics\RSquared;
 
 $estimator = new MLPRegressor([

docs/regressors/radius-neighbors-regressor.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/RadiusNeighborsRegressor.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/RadiusNeighborsRegressor/RadiusNeighborsRegressor.php">[source]</a></span>
 
 # Radius Neighbors Regressor
 This is the regressor version of [Radius Neighbors](../classifiers/radius-neighbors.md) implementing a binary spatial tree under the hood for fast radius queries. The prediction is a weighted average of each label from the training set that is within a fixed user-defined radius.
@@ -18,7 +18,7 @@ This is the regressor version of [Radius Neighbors](../classifiers/radius-neighb
 
 ## Example
 ```php
-use Rubix\ML\Regressors\RadiusNeighborsRegressor;
+use Rubix\ML\Regressors\RadiusNeighborsRegressor\RadiusNeighborsRegressor;
 use Rubix\ML\Graph\Trees\BallTree;
 use Rubix\ML\Kernels\Distance\Diagonal;
 

docs/regressors/regression-tree.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/RegressionTree.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/RegressionTree/RegressionTree.php">[source]</a></span>
 
 # Regression Tree
 A decision tree based on the CART (*Classification and Regression Tree*) learning algorithm that performs greedy splitting by minimizing the variance of the labels at each node split. Regression Trees can be used on their own or as the booster in algorithms such as [Gradient Boost](gradient-boost.md).
@@ -18,7 +18,7 @@ A decision tree based on the CART (*Classification and Regression Tree*) learnin
 
 ## Example
 ```php
-use Rubix\ML\Regressors\RegressionTree;
+use Rubix\ML\Regressors\RegressionTree\RegressionTree;
 
 $estimator = new RegressionTree(20, 2, 1e-3, 10, null);
 ```
@@ -50,4 +50,4 @@ public balance() : ?int
 
 ## References:
 [^1]: W. Y. Loh. (2011). Classification and Regression Trees.
-[^2]: K. Alsabti. et al. (1998). CLOUDS: A Decision Tree Classifier for Large Datasets.
+[^2]: K. Alsabti. et al. (1998). CLOUDS: A Decision Tree Classifier for Large Datasets.

docs/regressors/ridge.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/Ridge.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/Ridge/Ridge.php">[source]</a></span>
 
 # Ridge
 L2 regularized linear regression solved using a closed-form solution. The addition of regularization, controlled by the *alpha* hyper-parameter, makes Ridge less likely to overfit the training data than ordinary least squares (OLS).
@@ -14,7 +14,7 @@ L2 regularized linear regression solved using a closed-form solution. The additi
 
 ## Example
 ```php
-use Rubix\ML\Regressors\Ridge;
+use Rubix\ML\Regressors\Ridge\Ridge;
 
 $estimator = new Ridge(2.0);
 ```

docs/regressors/svr.md

@@ -1,4 +1,4 @@
-<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/SVR.php">[source]</a></span>
+<span style="float:right;"><a href="https://github.com/RubixML/ML/blob/master/src/Regressors/SVR/SVR.php">[source]</a></span>
 
 # SVR
 The Support Vector Machine Regressor (SVR) is a maximum margin algorithm for the purposes of regression. Similarly to the [SVC](../classifiers/svc.md), the model produced by SVR depends only on a subset of the training data, because the cost function for building the model ignores any training data close to the model prediction given by parameter *epsilon*. Thus, the value of epsilon defines a margin of tolerance where no penalty is given to errors.
@@ -33,7 +33,7 @@ public load(string $path) : void
 
 ## Example
 ```php
-use Rubix\ML\Regressors\SVR;
+use Rubix\ML\Regressors\SVR\SVR;
 use Rubix\ML\Kernels\SVM\RBF;
 
 $estimator = new SVR(1.0, 0.03, new RBF(), true, 1e-3, 256.0);