Changelog

1.2.1

Bug fixes:

Fix broken build.

1.2.0

Features:

When reading CSV files, assume columns in same order as defined in class if not given in file.
Add the Integer, Complex, Vector, and Tag feature types.
Add the Model.predict yield_inputs parameter.
Add the Model.train random_state parameter.

Bug fixes:

Support unicode in CSV files.
Fix bug which raised an error when predicting a Number that wasn't the first feature in Output.

1.1.0

Features:

Add basic feature type validation, and coercion, function.
Add test/train split parameters to Model.train.

Bug fixes:

Allow use of non-orderable types as Label labels.