James McCool
Enhance large_field_preset function to improve player exposure tracking by implementing a mechanism to remove high-exposure players across iterations, ensuring lineup integrity and preventing excessive ownership. Adjust logic to check for sufficient lineups before finalizing selections.
73d2fb0
| import pandas as pd | |
| def large_field_preset(portfolio: pd.DataFrame, lineup_target: int, exclude_cols: list): | |
| for slack_var in range(1, 20): | |
| concat_portfolio = pd.DataFrame(columns=portfolio.columns) | |
| # Define player columns (columns that contain player names) | |
| player_columns = [col for col in portfolio.columns if col not in exclude_cols] | |
| # Track players to remove across iterations | |
| remove_list = [] | |
| # Iterate until no high-exposure players are found | |
| max_iterations = 10 # Prevent infinite loops | |
| for iteration in range(max_iterations): | |
| concat_portfolio = pd.DataFrame(columns=portfolio.columns) | |
| for team in portfolio['Stack'].unique(): | |
| rows_to_drop = [] | |
| working_portfolio = portfolio.copy() | |
| # Remove players from previous iteration if any | |
| if remove_list: | |
| remove_mask = working_portfolio[player_columns].apply( | |
| lambda row: not any(player in list(row) for player in remove_list), axis=1 | |
| ) | |
| working_portfolio = working_portfolio[remove_mask] | |
| working_portfolio = working_portfolio[working_portfolio['Stack'] == team].sort_values(by='Finish_percentile', ascending = True) | |
| working_portfolio = working_portfolio.reset_index(drop=True) | |
| if len(working_portfolio) == 0: | |
| continue | |
| curr_own_type_max = working_portfolio.loc[0, 'Own'] + (slack_var / 20 * working_portfolio.loc[0, 'Own']) | |
| for i in range(1, len(working_portfolio)): | |
| if working_portfolio.loc[i, 'Own'] > curr_own_type_max: | |
| rows_to_drop.append(i) | |
| else: | |
| curr_own_type_max = working_portfolio.loc[i, 'Own'] + (slack_var / 20 * working_portfolio.loc[i, 'Own']) | |
| working_portfolio = working_portfolio.drop(rows_to_drop).reset_index(drop=True) | |
| concat_portfolio = pd.concat([concat_portfolio, working_portfolio]) | |
| # Check player exposure | |
| if len(concat_portfolio) == 0: | |
| break | |
| player_exposure = {} | |
| for col in player_columns: | |
| for player in concat_portfolio[col].unique(): | |
| if pd.notna(player): # Skip NaN values | |
| player_mask = concat_portfolio[player_columns].apply( | |
| lambda row: player in list(row), axis=1 | |
| ) | |
| exposure = player_mask.sum() / len(concat_portfolio) | |
| player_exposure[player] = exposure | |
| # Find players with exposure > 35% | |
| high_exposure_players = [player for player, exposure in player_exposure.items() if exposure > 0.35] | |
| # If no high-exposure players, we're done | |
| if not high_exposure_players: | |
| break | |
| # Add high-exposure players to remove list | |
| remove_list.extend(high_exposure_players) | |
| remove_list = list(set(remove_list)) # Remove duplicates | |
| # Check if we have enough lineups and no high-exposure players | |
| if len(concat_portfolio) >= lineup_target and len(high_exposure_players) == 0: | |
| return concat_portfolio.sort_values(by='Finish_percentile', ascending=True).head(lineup_target) | |
| return concat_portfolio.sort_values(by='Finish_percentile', ascending=True) | |