def test_asymmetric_rwlearner():
task = TwoArmedBandit()
critic = AsymmetricRescorlaWagnerLearner(task,
learning_rate_pos=0.1,
learning_rate_neg=0.1)
x = task.observation()
u = task.random_action()
x_, r, _ = task.step(u)
critic.update(x, u, r, x_, None)
def test_forgetful_rwlearner():
task = TwoArmedBandit()
critic = ForgetfulInstrumentalRescorlaWagnerLearner(task,
learning_rate=0.1,
memory_decay=0.1)
x = task.observation()
u = task.random_action()
x_, r, _ = task.step(u)
critic.update(x, u, r, x_, None)
def test_two_armed_bandit():
task = TwoArmedBandit()
x = task.observation()
u = task.random_action()