b15037e32ded46cc254d3bf3bcffc6ce8896a3ae,tests/algorithms/test_policy_gradient.py,,test_REINFORCE,#,49

Before Change

def test_REINFORCE():
    agent = REINFORCE(policy, mdp.info, **algorithm_params)
    agent.fit(dataset)

    w = np.array([-.07071067, 2.07071068])

    assert np.allclose(w, agent.policy.get_weights())

After Change

def test_REINFORCE():
    params = dict(learning_rate=AdaptiveParameter(value=.01))
    policy = learn(REINFORCE, params)
    w = np.array([-0.0084793 ,  2.00536528])

    assert np.allclose(w, policy.get_weights())

Instances