def visualize_b(state_action_partition):
vis_utils.plot_background(env, show=False)
vis_utils.plot_decision_boundary(g,
env.STATE_ACTION_MAP.shape[1],
env.STATE_ACTION_MAP.shape[0],
show=False)
vis_utils.plot_state_action_partition(state_action_partition,
show=True)
def visualize_b(state_action_partition):
vis_utils.plot_background(env, show=False)
xx, yy = np.meshgrid(np.arange(0, env.STATE_ACTION_MAP.shape[1], 0.01),
np.arange(0, env.STATE_ACTION_MAP.shape[0], 0.01))
data = np.c_[xx.ravel(), yy.ravel()]
Z = g.batch_predict(data[:, 0], data[:, 1])
Z = np.array(Z).reshape(xx.shape)
plt.contourf(xx, yy, Z, alpha=0.4)
vis_utils.plot_state_action_partition(state_action_partition, show=True)
def visualize_b(state_action_partition):
vis_utils.plot_background(env, show=False)
vis_utils.plot_state_action_partition(state_action_partition,
show=True)
def visualize_ignored(states):
vis_utils.plot_background(env, show=False)
plt.hist(states, bins=10, range=[0, len(env.STATE_MAP)], normed=True)
plt.show()