e8b2e17f4eec658f6a6d53486dbbe3eac48bb0fa,pyinterpret/core/global_interpretation/partial_dependence.py,PartialDependence,_plot_2d_1_binary_feature_and_1_continuous,#PartialDependence#Any#Any#Any#Any#Any#Any#,650

Before Change

            ax.set_ylabel(feature2)
            ax.set_zlabel("Predicted {}".format(class_name))
            handles, labels = ax.get_legend_handles_labels()
            ax.legend(handles, labels)
        return flatten([figure_list, axis_list])


    def partial_dependency_sklearn(self):

After Change

            ax = f.add_subplot(111)
            figure_list.append(f)
            axis_list.append(ax)
            plot_data = pdp.set_index([non_binary_feature, binary_feature])[mean_col]\
                .unstack()
            sd = pdp.set_index([non_binary_feature, binary_feature])[sd_col]\
                .unstack()

            plot_data.plot(ax=ax,color=COLORS)
            if with_variance:
                non_binary_values = plot_data.index.values
                binary_values = plot_data.columns.values
                upper_plane = plot_data + sd
                lower_plane = plot_data - sd
                for binary_value in binary_values:
                    color = next(colors)
                    ax.fill_between(non_binary_values, lower_plane[binary_value].values, upper_plane[binary_value].values, alpha=.2,
                                    color=color)

Instances