mirror of
https://github.com/wassname/pyrobolearn.git
synced 2026-09-20 13:11:01 +08:00
38 lines
1.0 KiB
Python
38 lines
1.0 KiB
Python
#!/usr/bin/env python
|
|
"""Example on how to use the 'Cartpole' OpenAI Gym environments in PyRoboLearn using a linear policy trained with
|
|
the Bayesian Optimization algorithm.
|
|
"""
|
|
|
|
import matplotlib.pyplot as plt
|
|
|
|
from pyrobolearn.envs import gym
|
|
from pyrobolearn.policies import LinearPolicy
|
|
from pyrobolearn.tasks import RLTask
|
|
from pyrobolearn.algos import BO
|
|
|
|
|
|
# create env, state, and action from gym
|
|
env = gym.make('CartPole-v1')
|
|
state, action = env.state, env.action
|
|
print("State and action space: {} and {}".format(state.space, action.space))
|
|
|
|
# create policy
|
|
policy = LinearPolicy(state, action)
|
|
|
|
# create task and run it
|
|
task = RLTask(env, policy)
|
|
task.run(num_steps=1000, use_terminating_condition=True, render=True)
|
|
|
|
# create RL algo
|
|
algo = BO(task, policy)
|
|
rewards = algo.train(num_steps=1000, num_episodes=30, verbose=True)
|
|
|
|
# plot
|
|
plt.figure()
|
|
plt.plot(rewards)
|
|
plt.show()
|
|
|
|
# test optimized policy
|
|
reward = algo.test(num_steps=1000, use_terminating_condition=True, render=True)
|
|
print("Final reward obtained on the test: {}".format(reward))
|