tests/testthat/test-decision_making.R

In konkam/ScopAI: This package plays at Scopa, and can learn to improve

test_that("Random decision works", {
  set.seed(1)
  for (i in 1:10) {
    g <- InitialiseGameState()
    decision <- RandomDecision(g, player = 1)
    expect_true(GetValueOfCard(decision$play) == GetSumValuesOfCards(decision$take) | is.null(decision$take))
  }

  game <- RunGame(seed = 1, starting_player = 1, DecisionFunction = RandomDecision)
  expect_false(any(is.na(game$score_player1)))
  expect_false(any(is.na(game$score_player2)))
  hands_player1 <- game$game_history %>% lapply(FUN = function(x) x$player1$hand)
  expect_equal(unique(hands_player1 %>% lapply(length)) %>% unlist() %>% sort(), 0:3)
  expect_false(any(hands_player1 %>% unlist() %>% is.na()))
  hands_player2 <- game$game_history %>% lapply(FUN = function(x) x$player2$hand)
  expect_equal(unique(hands_player2 %>% lapply(length)) %>% unlist() %>% sort(), 0:3)
  expect_false(any(hands_player2 %>% unlist() %>% is.na()))
})

test_that("Optimisable decision works", {
  set.seed(1)
  for (i in 1:10) {
    g <- InitialiseGameState()
    decision <- OptimisableDecision(g, player = 1, params = 0.5)
    expect_true(GetValueOfCard(decision$play) == GetSumValuesOfCards(decision$take) | is.null(decision$take))
  }
  OptimisedDecision <- function(g, player) OptimisableDecision(g, player, params = 0.5)
  game <- RunGame(seed = 1, starting_player = 1, DecisionFunction = OptimisedDecision)
  expect_false(any(is.na(game$score_player1)))
  expect_false(any(is.na(game$score_player2)))
  hands_player1 <- game$game_history %>% lapply(FUN = function(x) x$player1$hand)
  expect_equal(unique(hands_player1 %>% lapply(length)) %>% unlist() %>% sort(), 0:3)
  expect_false(any(hands_player1 %>% unlist() %>% is.na()))
  hands_player2 <- game$game_history %>% lapply(FUN = function(x) x$player2$hand)
  expect_equal(unique(hands_player2 %>% lapply(length)) %>% unlist() %>% sort(), 0:3)
  expect_false(any(hands_player2 %>% unlist() %>% is.na()))
})