Pyhsiclasso

The kernel to apply to the covariates can be changed (Gaussian or Delta).

Fixed covariate correction.

In addition, default version of HSIC Lasso is now the block version. Vanilla, non-block, HSIC Lasso can be used by setting block size B to 0.

The ndarray input function was not properly working.

1. The feature name index was stored as integer (this should be string).
2. We could not specify feature name.


v.1.3.5
Fixed Gaussian kernel width in multi-variate regressions.

- Fixed problems when dealing with singular matrices.

- Fixed ndarrays input not having featnames.

pyHSIC lasso speed up by ~30% by changing float64 to float32. Memory requirements halved.

release_v1.3.2
- Normalization is performed over the whole dataset, instead of on batches. This should result in a more robust normalization, and hence, better results.

release_v1.3.1
- Support parallel processing
- Multi-variate output support

release_v1.3.0
- Added Hierarchical Clustering
- Pull request 8
- Remove error when B is not a divisor of features
- Pull request 13
- accept discrete x
- Pull request 14

release_v1.2.0
Pull request 12

release_v1.1.0
Implemented block HSIC lasso
pull request 9