Benchmark autoplay strategies — benchmark_autoplay

Runs repeated seeded games for each autoplay strategy/settings row and returns both per-game runs and aggregate summaries. Parallel execution uses base R PSOCK clusters when workers > 1; sequential execution is the default for reproducible package checks.

Usage

benchmark_autoplay_strategies(
  config = game_config(),
  n_games = 50L,
  max_moves = 1000L,
  settings = autoplay_benchmark_settings("quality"),
  future_mode = c("visible", "rng"),
  seed = NULL,
  workers = 1L
)

Arguments

config: A configuration list created by game_config().
n_games: Number of games to run for each setting.
max_moves: Maximum moves per game.
settings: Strategy/settings grid from autoplay_benchmark_settings() or a compatible data frame.
future_mode: Future-tile handling passed to autoplay_move().
seed: Optional positive integer seed for deterministic game seeds. If NULL, a seed is drawn from R's current random-number stream.
workers: Number of PSOCK workers. Use 1 for sequential execution.

Value

A list with summary, runs, settings, config, and seed.