def test_does_not_modify_action(self):
env = PointEnv()
a = env.action_space.sample()
a_copy = a.copy()
env.reset()
env.step(a)
self.assertEquals(a.all(), a_copy.all())
def test_visualization(self):
env = PointEnv()
assert env.render_modes == ['ascii']
env.reset()
assert env.render('ascii') == f'Point: {env._point}, Goal: {env._goal}'
env.visualize()
env.step(env.action_space.sample())
def test_does_not_modify_action(self):
env = PointEnv()
a = env.action_space.sample()
a_copy = a.copy()
env.reset()
env.step(a)
assert a.all() == a_copy.all()
env.close()
def test_reset(self):
env = PointEnv()
env.reset()
assert (env._point == np.array([0, 0])).all()
a = env.action_space.sample()
env.step(a)
env.reset()
assert (env._point == np.array([0, 0])).all()
def test_observation_space(self):
env = PointEnv()
env.reset()
obs_space = env.observation_space
a = env.action_space.sample()
obs = env.step(a).observation
assert obs_space.contains(obs)
def test_done(self):
env = PointEnv()
for _ in range(1000):
_, _, done, _ = env.step(env._goal)
if done:
break
else:
assert False, 'Should report done'
def test_done(self):
env = PointEnv()
env.reset()
for _ in range(1000):
done = env.step(env._goal).terminal
if done:
break
else:
assert False, 'Should report done'
def test_catch_no_reset(self):
env = PointEnv()
with pytest.raises(RuntimeError, match='reset()'):
env.step(env.action_space.sample())