fixed pdp drop_na bug

RELEASE_NOTES.md

@@ -15,7 +15,7 @@
     stored parameters, e.g. `db = ExplainerDashboard.from_config("dashboard.yaml", higher_is_better=False)`
 
 ### Bug Fixes
--
+-   fixed bug where `drop_na=True` for `explainer.plot_pdp()` was not working.
 -
 
 ### Improvements

TODO.md

@@ -22,8 +22,6 @@
 
 
 ### Classifier plots:
-- add label percentage at cutoff to cumulative precision plot
-
 - add wizard to lift curve
 - pdp: add multiclass option
     - no icelines to keep it from getting too busy?
@@ -62,7 +60,6 @@
         - generate groups programmatically!
 
 ### Components
-- add description to PredictionSummaryComponent (reg+clas)
 - add hide_footer to components with a CardFooter
 
 - add querystring method to ExplainerComponents
@@ -88,8 +85,6 @@
 - Add this method? : https://arxiv.org/abs/2006.04750?
 
 ## Tests:
-- add prediction_summary_df test
-
 - test model_output='probability' and 'raw' or 'logodds' seperately
 - write tests for explainer_methods
 - write tests for explainer_plots

explainerdashboard/explainers.py

@@ -1525,7 +1525,7 @@ def plot_pdp(self, col, index=None, X_row=None, drop_na=True, sample=100,
 
         """
         pdp_result = self.get_pdp_result(col, index, X_row,
-                        num_grid_points=gridpoints, pos_label=pos_label)
+                        drop_na=drop_na, sample=sample, num_grid_points=gridpoints, pos_label=pos_label)
         units = "Predicted %" if self.model_output=='probability' else self.units
         if index is not None:
             col_value, pred = self.get_col_value_plus_prediction(col, index=index, pos_label=pos_label)