def test_state_encode_player():
env_state = MathyEnvState(problem="4x+2")
env_state = env_state.get_out_state(problem="2+4x",
moves_remaining=10,
action=(0, 0))
agent = env_state.agent
assert agent.problem == "2+4x"
assert agent.moves_remaining == 10
assert agent.action == (0, 0)
def test_state_serialize_numpy():
env_state = MathyEnvState(problem="4x+2")
for i in range(10):
env_state = env_state.get_out_state(problem="2+4x",
moves_remaining=10 - i,
action=(i, i))
state_np = env_state.to_np()
compare = MathyEnvState.from_np(state_np)
assert env_state.agent.problem == compare.agent.problem
assert env_state.agent.moves_remaining == compare.agent.moves_remaining
for one, two in zip(env_state.agent.history, compare.agent.history):
assert one.raw == two.raw
assert one.action == two.action