Libscientific

The fifth release release of libscientific.

**New features**

- No new features

**Bugfixes**

- Fixed important bug on LDA model
- Alignment of LDA model validation using the common implementation.

The fourth release of libscientific.

**New features**

- No new features

**Bugfixes**

- Fixed important bug on kfold_group_train_test_split.

The third release of libscientific.

**New features**

- New python bindings methods for matrix calculations
- Code readability improvement

**Bugfixes**

- Documentation moved to readthedocs.io. This close issues [2](https://github.com/gmrandazzo/libscientific/issues/2) and [3](https://github.com/gmrandazzo/libscientific/issues/3). Thank you to eriol and Beliavsky for the advices.
- Minor bug fixes on epls method.

Second release of libscientific

**New features**
- Python bindings
- Possibility to use the system lapack library
- Add support to handle handle missing values in matrix/vectors for PCA/PLS and other methods.

**Bugfixes**

- Add SONAME and Version control in CMAKE
- Documentation
- Readme
- Fixed the thread-safe random double generator
- Fixed important bug in MT_MatrixDVectorDotProduct
- Other minor bug fixes

This is the first release!

In this release:
- Fixed a memory bug in PLS algorithm
- Added PLS Discriminant Analyisis algorithm with binary tests (roc and precision-recall curves)
- Added generic model validation methods (KFold, Bootstrap, Leave One Out, Leave One group Out)
- Added Ensemble Partial Least Squares Regression and Discriminant Analysis using Boosting and Random Subspace Method