Method update

rl2/cartpole/q_learning_bins.py:68–70 · view source on GitHub ↗

(self, s, a, G)

Source from the content-addressed store, hash-verified

66	return self.Q[x]
67
68	def update(self, s, a, G):
69	x = self.feature_transformer.transform(s)
70	self.Q[x,a] += 1e-2*(G - self.Q[x,a])
71
72	def sample_action(self, s, eps):
73	if np.random.random() < eps:

play_oneFunction · 0.45

transformMethod · 0.45

no test coverage detected