v0.1.3 (September 23, 2022)

This is a minor release with a change in objective and bug fixes for issues identified since v0.1.2.

Features

• Change made in the software orientation by starting developing the packages following the scikit-learn API. The fit method is implemented in many functions and replaces the GMT development API. This has happened when the models module is starting growing by implementing the cross-validation.

• New algorithmns in the watex.base module are started such as watex.base.SequentialBackwardSelection, watex.base.AdalineGradientDescent additional to the assessors watex.base.Data and watex.base.Missing.

• Upload the watex documentation in ReadTheDocs server for the first time. See older v0.1.3 documentation

• Upload the _p module ( from watex.datasets._p) referring to an example of data preparation, the case of Bagoue Region.

• Upload new packages of datasets such watex.datasets.load_semien(), watex.datasets.load_boundiali(), watex.datasets.load_gbalo(), watex.datasets.load_iris().

Bug fixes

• Fixed a bug in watex.view.QuickPlot when setting objective to flow and expecting the data to be automated following the case-history paper of Bagoue region.

• Fixed a bug in fetching data in remote from watex.datasets.rload.loadBagoueDataset(). Unable to fetch correctly the dataset reset every time a new data record is uploaded in zenodo.

• Fixed bug in watex.datasets.rload.Loader by mixing the three ways of fetching data from _fromgithub, _fromzenodo, _fromlocal to optmize the change of recovering data. A thread is set for parallelizing the data retrieval.