from pybrain.rl.environments.cartpole import CartPoleEnvironment, CartPoleRenderer, BalanceTask
from pybrain.rl.agents.learning import LearningAgent
from pybrain.rl.experiments import EpisodicExperiment
from scipy import mean
import sys
episodes = 1
epilen = 200
if len(sys.argv) < 5:
sys.exit('please give 4 parameters. run: "python play_catpole.py <p1> <p2> <p3> <p4>"\n')
# create environment
env = CartPoleEnvironment()
env.setRenderer(CartPoleRenderer())
env.getRenderer().start()
env.delay = (episodes == 1)
# create task
task = BalanceTask(env, epilen)
# create controller network
net = buildNetwork(4, 1, bias=False)
# create agent and set parameters from command line
agent = LearningAgent(net, None)
agent.module._setParameters([float(sys.argv[1]), float(sys.argv[2]), float(sys.argv[3]), float(sys.argv[4])])
# create experiment
experiment = EpisodicExperiment(task, agent)
experiment.doEpisodes(episodes)
from pybrain.rl.agents.learning import LearningAgent
from pybrain.rl.experiments import EpisodicExperiment
from scipy import mean
import sys
import pylab
episodes = 1
epilen = 200
if len(sys.argv) < 5:
sys.exit('please give 4 parameters. run: "python play_catpole.py <p1> <p2> <p3> <p4>"\n')
# create environment
env = CartPoleEnvironment()
env.setRenderer(CartPoleRenderer())
env.getRenderer().start()
env.delay = (episodes == 1)
# create task
task = BalanceTask(env, epilen)
# create controller network
net = buildNetwork(4, 1, bias=False)
# create agent and set parameters from command line
agent = LearningAgent(net, None)
arg1 = float(sys.argv[1])
arg2 = float(sys.argv[2])
arg3 = float(sys.argv[3])
arg4 = float(sys.argv[4])
agent.module._setParameters([arg1, arg2, arg3, arg4])
