GitLab

* Update to phpunit 8

* Require at least PHP 7.2

* Add NGramWordTokenizer

* Update doc
Add test
Check coding standards

* Issue #349: Provide a new NGramTokenizer.

* Issue #349: Add tests.

* Fixes from code review.

* Implement NGramTokenizer with min and max gram support

* Add missing tests for ngram

* Add info about NGramTokenizer to docs and readme

* Add performance test for tokenization

* Issue #347: Reduce duplicated code.

* Issue #347: Replace array_* with regular loops for better perfomance.

* Update phpunit to 7.5.1

* Remove osx build from travis

* Check if feature exist when predict target in NaiveBayes

* Fix typo

* Implement MnistDataset

* Add MNIST dataset documentation

* Update .travis.yml

* remove coverage-clover for 7.2

* move coverage-clover to 7.2

* Update docs for Imputer class

* Throw exception when trying to transform imputer without train data

* Update changelog

* Change the default kernel type in SVC to Kernel::RBF

* Update CHANGELOG.md

* update ECS config to v4

* composer: require Symplify 4

* apply coding-standard: use constants over functions, protected setUp() in tests, array indentation

* ecs: add false positive case

* composer: update lock

* bump to ECS 4.4

* update composer.lock

* shorten ECS config name

* ecs: ignore assignments in while()

* fix cs

* Update osx build for travis

* Try something more with this os

* This gonna be hard ...

* Cleanup travis build even more

* Install phpbench 🚀

* Add first benchmark for LeastSquares

* Update README and CONTRIBUTING guide

* Fix typo

* KMeans associative clustering added

* fix travis error

* KMeans will return provided keys as point label if they are provided

* fix travis

* fix travis

* Micro optimization for matrix multiplication

* code cs fix

* added a comment block for the change

* Throw proper exception

* Fix coding style

* Add  Optimizer tests

* Remove Optimizer.initialTheta and rename Optimizer.setInitialTheta to setTheta