best_15_optimisation module

Source file that holds the model formulation of the best 15 players optimisation problem.

Functions in the source file:
exception best_15_optimisation.OptimisationValuesAllZeroError

Bases: Exception

Exception for when the values chosen to be used as the main optimisation values in find_best_15_players_by_value() are all zero. This would technically happen only in the pre-season period when for example ‘Form’ is zero for all players.

best_15_optimisation.find_best_15_players_by_value(player_ids: List, player_positions: List, player_values: List, player_prices: List, player_teams: List, opt_target: str, players_pre_selected: List = None)

Calculates the best 15 player selection according to the value passed as an argument. Uses the PULP library and default CBC solver. It satisfies the max 3 players per team constraint and the 100 cost constraint.

Parameters:

  • player_ids (list) – Player unique ids

  • player_positions (list) – Player positions

  • player_values (list) – Player values

  • player_prices (list) – Player prices

  • player_teams (list) – Player teams

  • opt_target (str) – optimisation target (the target value)

  • players_pre_selected (list) – Players pre-selected by the user (forced to be included)

Returns:

two pandas dataframes, first containing the players and their details, the second the optimisation information

best_15_optimisation.post_process_data(players_df: DataFrame, results: DataFrame, statistics: DataFrame)

Separate the results returned by the optimisation and bring it to a format suitable for display.

Parameters:

  • players_df (pandas.dataframe) – Dataframe containing the uid, first and last name of players

  • results (pandas.dataframe) – The results of the optimisation

  • statistics (pandas.dataframe) – pandas dataframe containing the statistics of the optimisation process