Refactor reassess_finish_percentile function in reassess_edge.py to simplify calculations by using ownership and median differences, improving accuracy in finish_percentile adjustments. Update column drop logic to remove unnecessary columns for cleaner data handling.

global_func/reassess_edge.py

@@ -44,31 +44,10 @@ def calculate_weighted_ownership_single_row(row_ownerships):
     # Convert back to percentage form
     return weighted * 10000
 
-def reassess_finish_percentile(row: pd.Series, own_columns: list, Contest_Size: int) -> float:
-    """
-    Recalculate Finish_percentile for an individual row.
-    
-    Args:
-        row: pandas Series containing the row data
-        own_columns: list of ownership column names
-        
-    Returns:
-        float: new finish_percentile value
-    """
-    percentile_multiplier = 1.00
-    
-    # Step 1: Adjust Finish_percentile
-    finish_percentile = row['Finish_percentile'] + 0.005 + (0.005 * (Contest_Size / 10000))
-    
-    # Step 2: Apply multiplier and ownership adjustment
-    finish_percentile = finish_percentile * percentile_multiplier * (row['Own'] / (100 + (Contest_Size / 1000)))
-    
-    # Step 3: Count low ownership players
-    low_own_count = sum(1 for col in own_columns if row[col] < 0.10)
-    
-    # Step 4: Apply low ownership adjustment
-    if low_own_count > 0:
-        finish_percentile = finish_percentile / low_own_count
+def reassess_finish_percentile(row: pd.Series, own_columns: list) -> float:
+    own_diff = float(row['own_diff'])
+    median_diff = float(row['median_diff'])
+    finish_percentile = row['Finish_percentile'] + (own_diff / 2) - (median_diff / 100)
     
     return finish_percentile
 
@@ -139,6 +118,6 @@ def reassess_edge(refactored_frame: pd.DataFrame, original_frame: pd.DataFrame,
         refactored_df.loc[lineups, 'Weighted Own'] = calculate_weighted_ownership_single_row(refactored_df.loc[lineups, own_columns])
         refactored_df.loc[lineups, 'Geomean'] = np.power((refactored_df.loc[lineups, own_columns] * 100).product(), 1 / len(own_columns))
     
-    refactored_df = refactored_df.drop(columns=own_columns)
+    refactored_df = refactored_df.drop(columns=['salary_diff', 'median_diff', 'own_diff', 'player_0_own', 'Edge'] + own_columns)
 
     return refactored_df