- extended documentation (cheatsheet)
- changed the way plot(model_triplot) is build, so it works correctly with patchwork 1.0.1
- changed parameter in DALEX::feature_importance call from n_sample to N (due to change in DALEX 1.2.1)
- added parameter to change correlation method in group_variables, cluster_variables, triplot
- deprecated aspects_on_axis argument in plot.aspect_importance (now it’s show_features) (#14)
- last vertical line in hierarchical_importance() plot of model_triplot() shows baseline value
- left margin of middle panel in triplot can be set with "margin_mid"
- colors in plot for predict_aspects() changed to match DALEX::colors_discrete_drwhy()
- changed order or n_var and sample_method parameters in aspect_importance() functions
- renamed triplot title to Global / Local triplot
- modified triplot object and methods
- removed ingredients dependency
- switched from GridExtra to Patchwork
- improved plotting of triplot object
- renamed prediction_aspect() to predict_aspects()
- added aliases predict_triplot() and model_triplot()
- added bar_width param to triplot
- added warning if target is in data
- added parameter "type" to calculate_triplot()
- added print for predict_aspects() and triplot()
- restructured group_variables() and triplot objects
- added hierarchical_importance object (instead of plot_aspects_importance_grouping)
- triplot renamed to calculate_triplot()
- triplot is plotted by calling plot on calculate_triplot() results
- updated plot_aspects_importance_grouping so it can draw last clusters connection
- add_importance_labels parameter set to FALSE by default
- new feature: triplot can present both: automatic aspect and feature importance grouping (by using ingredients::feature_importance())
- aspectImportance is renamed to triplot
- added second vignette
- aspect_importance is moved out of
DALEXtrainto standalone package