Update app.py

app.py

@@ -64,22 +64,21 @@ def calcular_fator_avaliacao(titulo, EC, PU):
     fator_avaliacao = max((4 * ec_pontuacao + 6 * PVU - 3 * PUB) / 100, VR)
     return fator_avaliacao
 
-def select_nearest_items(df, n=5):
+def select_nearest_items(df):
     median_price = df['Price'].median()
     df['Distance'] = (df['Price'] - median_price).abs()
-    df_sorted = df.sort_values(by=['Distance', 'Marketplace'])  # Sort by distance then by marketplace to diversify
-    selected_items = pd.DataFrame()
-
-    marketplaces_seen = set()
+    df_sorted = df.sort_values('Distance')
+    
+    # Ensuring unique marketplaces, maintaining a list of included marketplaces
+    included_marketplaces = set()
+    nearest_items = pd.DataFrame()
     for _, row in df_sorted.iterrows():
-        if len(selected_items) >= n:
+        if row['Marketplace'] not in included_marketplaces:
+            nearest_items = nearest_items.append(row, ignore_index=True)
+            included_marketplaces.add(row['Marketplace'])
+        if len(included_marketplaces) == 5:
             break
-        if row['Marketplace'] not in marketplaces_seen:
-            selected_items = selected_items.append(row, ignore_index=True)
-            marketplaces_seen.add(row['Marketplace'])
-
-    selected_items = selected_items.drop(columns=['Distance'])  # Remove the temporary 'Distance' column
-    return selected_items
+    return nearest_items
 
 def integrated_app(query, titulo, EC, PU):
     df_mercadolibre = fetch_data_to_dataframe(query, 50, "mercadolibre")