switched ranking of scores to "first" meaning that equal entities get…

… a different rank based on the order in the input array
dssg · Apr 27, 2017 · 93cc818 · 93cc818
1 parent 4144690
commit 93cc818
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 4 deletions.
diff --git a/triage/model_trainers.py b/triage/model_trainers.py
@@ -164,8 +164,8 @@ def _write_model_to_db(
         feature_importance = get_feature_importances(trained_model)
         temp_df = pandas.DataFrame({'feature_importance': feature_importance})
         features_index = temp_df.index.tolist()
-        rankings_abs = temp_df['feature_importance'].rank(method='dense', ascending=False)
-        rankings_pct = temp_df['feature_importance'].rank(method='dense', ascending=False, pct=True)
+        rankings_abs = temp_df['feature_importance'].rank(method='first', ascending=False)
+        rankings_pct = temp_df['feature_importance'].rank(method='first', ascending=False, pct=True)
         for feature_index, importance, rank_abs, rank_pct in zip(
             features_index,
             feature_importance,

diff --git a/triage/predictors.py b/triage/predictors.py
@@ -100,8 +100,8 @@ def _write_to_db(
                 ))
         else:
             temp_df = pandas.DataFrame({'score': predictions})
-            rankings_abs = temp_df['score'].rank(method='dense', ascending=False)
-            rankings_pct = temp_df['score'].rank(method='dense', ascending=False, pct=True)
+            rankings_abs = temp_df['score'].rank(method='first', ascending=False)
+            rankings_pct = temp_df['score'].rank(method='first', ascending=False, pct=True)
             for entity_id, score, label, rank_abs, rank_pct in zip(
                 matrix.index,
                 predictions,