Fix and unit test for issue "Wrong style for state report for TicTacT…

…oeEnv() JuliaReinforcementLearning#1079"
hespanha · Sep 3, 2024 · 43a13a4 · 43a13a4
1 parent b49bf69
commit 43a13a4
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 4 deletions.
diff --git a/src/ReinforcementLearningEnvironments/src/environments/examples/TicTacToeEnv.jl b/src/ReinforcementLearningEnvironments/src/environments/examples/TicTacToeEnv.jl
@@ -72,6 +72,7 @@ RLBase.players(::TicTacToeEnv) = (Player(:Cross), Player(:Nought))
 
 RLBase.state(env::TicTacToeEnv, ::Observation, ::DefaultPlayer) = state(env, Observation{Int}(), Player(:Any))
 RLBase.state(env::TicTacToeEnv, ::Observation{BitArray{3}}, player) = env.board
+RLBase.state(env::TicTacToeEnv, ::Observation{BitArray{3}}) = env.board
 RLBase.state(env::TicTacToeEnv, ::RLBase.AbstractStateStyle) = state(env::TicTacToeEnv, Observation{Int}(), Player(1))
 RLBase.state(env::TicTacToeEnv, ::Observation{Int}, player::Player) =
     get_tic_tac_toe_state_info()[env].index

diff --git a/src/ReinforcementLearningEnvironments/test/environments/examples/tic_tac_toe.jl b/src/ReinforcementLearningEnvironments/test/environments/examples/tic_tac_toe.jl
@@ -3,15 +3,15 @@
     using ReinforcementLearningEnvironments, ReinforcementLearningBase, ReinforcementLearningCore
 
     trajectory_1 = Trajectory(
-        CircularArraySARTSTraces(; capacity = 1),
+        CircularArraySARTSTraces(; capacity=1),
         BatchSampler(1),
-        InsertSampleRatioController(n_inserted = -1),
+        InsertSampleRatioController(n_inserted=-1),
     )
 
     trajectory_2 = Trajectory(
-        CircularArraySARTSTraces(; capacity = 1),
+        CircularArraySARTSTraces(; capacity=1),
         BatchSampler(1),
-        InsertSampleRatioController(n_inserted = -1),
+        InsertSampleRatioController(n_inserted=-1),
     )
 
     multiagent_policy = MultiAgentPolicy(PlayerTuple(
@@ -30,6 +30,7 @@
     @test length(state_space(env, Observation{Int}())) == 5478
 
     @test RLBase.state(env, Observation{BitArray{3}}(), Player(:Cross)) == env.board
+    @test RLBase.state(env, Observation{BitArray{3}}()) == env.board
     @test RLBase.state_space(env, Observation{BitArray{3}}(), Player(:Cross)) isa ArrayProductDomain
     @test RLBase.state_space(env, Observation{String}(), Player(:Cross)) isa DomainSets.FullSpace{String}
     @test RLBase.state(env, Observation{String}(), Player(:Cross)) isa String