3faa7a095c740b227f16483c4d17721bfb2bb17a,dm_control/suite/tests/domains_test.py,DomainTest,test_visualize_reward,#DomainTest#Any#Any#,188

Before Change

    env.task.visualise_reward = True
    env.reset()
    action = np.zeros(env.action_spec().shape)
    for _ in range(2):
      env.step(action)

  @parameterized.parameters(*suite.ALL_TASKS)
  def test_task_supports_environment_kwargs(self, domain, task):
    env = suite.load(domain, task,
                     environment_kwargs=dict(flat_observation=True))

After Change

      mock_get_reward.reset_mock()
      mock_get_reward.return_value = 2.0  // Rewards > 1 should be clipped.
      env.step(action)
      mock_get_reward.assert_called_with(env.physics)
      self.assertCorrectColors(env.physics, reward=1.0)

      mock_get_reward.reset_mock()

Instances