v1.4.0

Highlights

The binning methods of numpy.histogram_bin_edges are now available in compute_bias, compute_marginal, plot_bias and plot_marginal.
The default argument of bin_method of these functions changed to "auto".
Additionally, strings/categorical/enum features now display all the remaining (unique) values of a feature as "rest-n" where n indicates the number of unique values. This way, n_bins is strictly adhered to.

The marginal plot from the example Regression on Workers' Compensation Dataset now looks like this:

What's Changed

• MNT hatchling >= 1.26.3 by @lorentzenchr in #192
• DOC improve docstring of add_marginal_subplot by @lorentzenchr in #193
• ENH add numpy binning options by @lorentzenchr in #194
• REL increase to version 1.4.0 by @lorentzenchr in #195

Full Changelog: v1.3.0...v1.4.0