fix get-item-based-k

azureml-designer-modules/entries/score_sar_entry.py

@@ -40,7 +40,7 @@ def recommend_items(model, data, ranking_metric, top_k, sort_top_k, remove_seen,
     raise ValueError(f"Got unexpected ranking metric: {ranking_metric}.")
 
 
-def predict_ratings(model, data, items_to_predict, remove_seen, normalize):
+def predict_ratings(model, data, items_to_predict, normalize):
     if items_to_predict == ItemSet.TRAIN_ONLY:
         return model.predict_training_items(test=data, normalize=normalize)
     if items_to_predict == ItemSet.SCORE_ONLY:
@@ -101,7 +101,6 @@ def predict_ratings(model, data, items_to_predict, remove_seen, normalize):
         score_result = predict_ratings(model=sar_model,
                                        data=dataset_to_score,
                                        items_to_predict=ItemSet(args.items_to_predict),
-                                       remove_seen=remove_seen_items,
                                        normalize=normalize)
     else:
         raise ValueError(f"Got unexpected score type: {score_type}.")

azureml-designer-modules/module_specs/sar_score.yaml

@@ -1,5 +1,5 @@
 name: SAR Score
-version: 0.0.19
+version: 0.0.20
 category: Experimentation
 description: |
     Python SAR Recommender

reco_utils/recommender/sar/sar_singlenode.py

@@ -367,7 +367,13 @@ def get_item_based_topk(self, items, top_k=10, sort_top_k=True):
         """
 
         # convert item ids to indices
-        item_ids = items[self.col_item].map(self.item2index)
+        items[self.col_item] = items[self.col_item].map(self.item2index)
+        if any(items.isna()):
+            logger.warning("Items found in test not seen during training, new items won't be considered as "
+                           "seed item of recommendations.")
+            items = items.dropna()
+
+        item_ids = items[self.col_item].astype(int)
 
         # if no ratings were provided assume they are all 1
         if self.col_rating in items.columns: