Struct open_ttt_lib::ai::Opponent

pub struct Opponent { /* fields omitted */ }

Provides a computer controlled AI opponent.

This can be used to create single player games or implement a hint system for human users.

Methods

impl Opponent

pub fn new(mistake_probability: f64) -> Self

Constructs a new AI opponent.

The mistake probability indicates how likely the AI will fail to consider various situations. A value of 0.0 makes the AI play a perfect game. A value of 1.0 causes the AI to always pick a random position. Values less than 0.0 are set to 0.0 and values greater than 1.0 are set to 1.0.

Examples

Construct an unbeatable AI opponent:

use open_ttt_lib::ai;

let mistake_probability = 0.0;
let unbeatable_opponent = ai::Opponent::new(mistake_probability);

Construct an AI opponent that randomly picks a position:

use open_ttt_lib::ai;

let mistake_probability = 1.0;
let rando = ai::Opponent::new(mistake_probability);

pub fn get_move(&self, game: &Game) -> Option<Position>

Gets the position the AI opponent wishes to move based on the provided game.

None is returned if the game is over. The AI opponent never tries to select an invalid position, that is a position that is not free.

Examples

use open_ttt_lib::ai;
use open_ttt_lib::game;

let game = game::Game::new();
let ai_opponent = ai::Opponent::new(0.0);

match ai_opponent.get_move(&game) {
    Some(position) => assert!(game.can_move(position)),
    None => panic!("The game is over so the AI opponent cannot do a move."),
};

pub fn evaluate_game(&self, game: &Game) -> HashMap<Position, Outcome>

Evaluates each free position in the provided game.

Each free position in the game is mapped to an outcome for the AI opponent. If the game is over an empty map is returned.

This functionality is useful if you wish to examine how the AI opponent viewed the game. E.g. this can be helpful for creating a hint system to help human players pick a position or when fine tuning the AI difficulty settings.

Examples

use open_ttt_lib::ai;
use open_ttt_lib::game;

let game = game::Game::new();
let ai_opponent = ai::Opponent::new(0.0);

let outcomes = ai_opponent.evaluate_game(&game);

// Display the outcome for each position.
for (position, outcome) in outcomes {
    assert!(game.can_move(position));
    println!("position: {:?} outcome: {:?}", position, outcome);
}