Update app.py

app.py

@@ -598,19 +598,22 @@ def server(input, output, session):
 
 
 
-        # List of desired columns
+        # List of desired columns, including RHP and LHP
         desired_columns = [
             'Player ID', 'Batter', 'Bats', 'GP', '1', '2', '3', '4', '5', '6', '7', '8', '9',
             'C', '1B', '2B', '3B', 'SS', 'LF', 'CF', 'RF', 'DH', 'LHP', 'RHP'
         ]
         
-        # Check for existing columns and select safely
-        existing_columns = [col for col in desired_columns if col in df_test_merge.columns]
+        # Ensure all desired columns exist in the Polars DataFrame.
+        # If a column is missing, add it with a default value (e.g., empty string '')
+        for col in desired_columns:
+            if col not in df_test_merge.columns:
+                df_test_merge = df_test_merge.with_columns(pl.lit('').alias(col))
         
-        # Select existing columns
-        df_test_merge = df_test_merge.select(existing_columns)
+        # Now select the columns in the desired order
+        df_test_merge = df_test_merge.select(desired_columns)
         
-        # Continue with the rest of the code
+        # Define your mapping for hierarchical columns.
         cols = {
             ('Player Info', 'Player ID'): 'Player ID',
             ('Player Info', 'Batter'): 'Batter',
@@ -638,14 +641,32 @@ def server(input, output, session):
             ('Pitcher', 'RHP'): 'RHP'
         }
         
-        # Convert to pandas
+        # Convert the Polars DataFrame to a pandas DataFrame
         df_pandas = df_test_merge.to_pandas()
         
-        # Replace 0 with ''
+        # Replace any zeros with empty strings (if needed)
         df_pandas = df_pandas.replace({0: ''})
         
-        # Rename columns with multi-index
-        df_pandas.columns = pd.MultiIndex.from_tuples([(k[0], k[1]) for k in cols.keys()])
+        # Create the MultiIndex for columns.
+        # The order of tuples should correspond exactly to the order in desired_columns.
+        # We can build the new MultiIndex by mapping the desired_columns using the `cols` dictionary.
+        new_columns = []
+        for col in desired_columns:
+            # Find the key in cols where value matches col.
+            # This assumes your mapping has unique values.
+            for key, value in cols.items():
+                if value == col:
+                    new_columns.append(key)
+                    break
+        
+        # Double-check that the length of new_columns matches the DataFrame columns.
+        if len(new_columns) != len(df_pandas.columns):
+            raise ValueError("Mismatch in number of columns between new MultiIndex and DataFrame columns.")
+        
+        # Apply the new MultiIndex to the DataFrame
+        df_pandas.columns = pd.MultiIndex.from_tuples(new_columns)
+        
+        # Now your DataFrame has hierarchical columns and guaranteed 'LHP' and 'RHP' columns.
 
         df_pivot_style = df_pandas.style
         thick_border_cols = [3, 4, 13,22]  # 0-based index