Reference

class poke_engine.IterativeDeepeningResult(side_one: list[str], side_two: list[str], matrix: list[float], depth_searched: int)

Result of an Iterative Deepening Expectiminimax Search

Parameters:

  • side_one (list[str]) – The moves for side_one

  • side_two (list[str]) – The moves for side_two

  • matrix (int) – A vector representing the payoff matrix of the search. Pruned branches are represented by None

  • depth_searched (int) – The depth that was searched to

get_safest_move() → str

Get the safest move for side_one The safest move is the move that minimizes the loss for the turn

Returns:

The safest move

Return type:

str

class poke_engine.MctsResult(side_one: list[MctsSideResult], side_two: list[MctsSideResult], total_visits: int)

Result of a Monte Carlo Tree Search

Parameters:

  • side_one (list[MctsSideResult]) – Result for side one

  • side_two (list[MctsSideResult]) – Result for side two

  • total_visits (int) – Total number of monte carlo iterations

class poke_engine.MctsSideResult(move_choice: str, total_score: float, visits: int)

Result of a Monte Carlo Tree Search for a single side

Parameters:

  • move_choice (str) – The move that was chosen

  • total_score (float) – The total score of the chosen move

  • visits (int) – The number of times the move was chosen

class poke_engine.PokemonIndex(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
class poke_engine.Terrain(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
class poke_engine.Weather(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)
poke_engine.iterative_deepening_expectiminimax(state: State, duration_ms: int = 1000) → IterativeDeepeningResult

Perform an iterative-deepening expectiminimax search on the given state and for the given duration

Parameters:

  • state (State) – the state to search through

  • duration_ms (int) – time in milliseconds to run the search

Returns:

the result of the search

Return type:

IterativeDeepeningResult

Perform monte-carlo-tree-search on the given state and for the given duration

Parameters:

  • state (State) – the state to search through

  • duration_ms (int) – time in milliseconds to run the search

Returns:

the result of the search

Return type:

MctsResult