Debugging

DF_Construction.ipynb

@@ -2899,7 +2899,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 39,
    "id": "05811f9a",
    "metadata": {},
    "outputs": [],
@@ -2929,7 +2929,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 40,
    "id": "78eb378f",
    "metadata": {},
    "outputs": [
@@ -2939,7 +2939,7 @@
        "42"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 40,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2953,7 +2953,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 42,
    "id": "b7a07409",
    "metadata": {},
    "outputs": [
@@ -2982,7 +2982,7 @@
     "    filename = \"users_{}.csv\".format(part)\n",
     "    start_idx = part * n_rows\n",
     "    end_idx = min((part+1) * n_rows, users_df.shape[0])\n",
-    "    users_df.loc[start_idx:start_idx].to_csv(filename, index=None)"
+    "    users_df.loc[start_idx:end_idx].to_csv(filename, index=None)"
    ]
   },
   {
@@ -2996,7 +2996,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 35,
    "id": "101ade69",
    "metadata": {},
    "outputs": [
@@ -3006,7 +3006,7 @@
        "9015"
       ]
      },
-     "execution_count": 11,
+     "execution_count": 35,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -3020,7 +3020,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 36,
    "id": "35266a30",
    "metadata": {},
    "outputs": [
@@ -3042,7 +3042,7 @@
     "    filename = \"items_{}.csv\".format(part)\n",
     "    start_idx = part * n_rows\n",
     "    end_idx = min((part+1) * n_rows, items_df.shape[0])\n",
-    "    items_df.loc[start_idx:start_idx].to_csv(filename, index=None)"
+    "    items_df.loc[start_idx:end_idx].to_csv(filename, index=None)"
    ]
   },
   {
@@ -3185,7 +3185,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 22,
    "id": "ac68d277",
    "metadata": {},
    "outputs": [
@@ -3307,7 +3307,7 @@
        "[2 rows x 30 columns]"
       ]
      },
-     "execution_count": 21,
+     "execution_count": 22,
      "metadata": {},
      "output_type": "execute_result"
     }

app.py

@@ -24,7 +24,7 @@ with gr.Blocks(css=css) as demo:
 
       # First Component
       with gr.Column(scale=1):
-        col_1_number = gr.Number(value=1, visible=False)
+        col_1_number = gr.Number(value=0, visible=False)
         
         gr.Markdown("Based on similar movies...")
 
@@ -41,7 +41,7 @@ with gr.Blocks(css=css) as demo:
 
       # Second Component
       with gr.Column(scale=1):
-        col_2_number = gr.Number(value=2, visible=False)
+        col_2_number = gr.Number(value=1, visible=False)
         
         gr.Markdown("Based on similar users...")
 
@@ -58,7 +58,7 @@ with gr.Blocks(css=css) as demo:
 
       # Third Component
       with gr.Column(scale=1):
-        col_3_number = gr.Number(value=3, visible=False)
+        col_3_number = gr.Number(value=2, visible=False)
 
         gr.Markdown("Based on similar users and movies...")
 

recommender.py

@@ -49,7 +49,7 @@ class ItemData:
         return self.df[self.df['id'] == int(id)][self.item_columns]
     
     def get_id_by_idx(self, idx):
-        return self.df.at[idx, 'id']
+        return self.df.iloc[int(idx)]['id']
     
     def get_random_id(self):
         return self.df.sample(1)['id'].values[0]
@@ -181,7 +181,7 @@ class Recommender:
                     cols_to_drop.append(best_movie)
                     continue
                 
-                if self.user_handler.df.at[best_idx, best_movie] > 2.5:
+                if self.user_handler.df.iloc[best_idx][best_movie] > 2.5:
                     return int(best_movie)
 
         # Otherwise, return random movie
@@ -216,7 +216,7 @@ class Recommender:
                 continue
 
             for user_id in user_idxs[0]:
-                score = self.user_handler.df.at[user_id, str(movie_id)]
+                score = self.user_handler.df.iloc[user_id][str(movie_id)]
                 if score != 0:
                     score_sums[i] += score
                     n_votes[i] += 1
@@ -233,4 +233,19 @@ class Recommender:
         # Find best score
         best_score_idx = final_score.index(max(final_score))
         best_movie_idx = item_idxs[0][best_score_idx]
-        return self.item_handler.get_id_by_idx(best_movie_idx)
+        return self.item_handler.get_id_by_idx(best_movie_idx)
+
+
+if __name__ == '__main__':
+    recommender = Recommender()
+    res = recommender.on_pick(0, 1)
+    print(res)
+    res = recommender.on_pick(0, 3.4)
+    print(res)
+    res = recommender.on_pick(1, 1)
+    print(res)
+    res = recommender.on_pick(1, 3.4)
+    print(res)
+    
+    # print(recommender.item_handler.df.iloc[20712]['id'])
+