本文整理汇总了Python中pybrain.rl.agents.LearningAgent.name方法的典型用法代码示例。如果您正苦于以下问题：Python LearningAgent.name方法的具体用法？Python LearningAgent.name怎么用？Python LearningAgent.name使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类pybrain.rl.agents.LearningAgent的用法示例。
在下文中一共展示了LearningAgent.name方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: ENAC
# 需要导入模块: from pybrain.rl.agents import LearningAgent [as 别名]
# 或者: from pybrain.rl.agents.LearningAgent import name [as 别名]
# Create an agent and select an episodic learner.
# learner = ENAC()
learner = Reinforce()
learner.gd.rprop = False
# only relevant for BP
# learner.learningRate = 0.001 # (0.1-0.001, down to 1e-7 for RNNs, default: 0.1)
learner.gd.alpha = 0.01
# learner.gd.alphadecay = 0.9
# learner.gd.momentum = 0.9
# only relevant for RP
# learner.gd.deltamin = 0.0001
agent = LearningAgent(net, learner)
# Name the agent according to its first generator's name.
agent.name = gen.name
# Adjust some parameters of the NormalExplorer.
if manual_sigma:
sigma = [-5.0] * env.indim
learner.explorer.sigma = sigma
# Add the task and agent to the experiment.
experiment.tasks.append(task)
experiment.agents.append(agent)
takers = case.generators[1:]
for g in takers:
env = pyreto.continuous.MarketEnvironment([g], market, numOffbids)
task = pyreto.continuous.ProfitTask(env, maxSteps=len(p1h))
agent = pyreto.util.NegOneAgent(env.outdim, env.indim)
experiment.tasks.append(task)示例2: MarketExperiment
# 需要导入模块: from pybrain.rl.agents import LearningAgent [as 别名]
# 或者: from pybrain.rl.agents.LearningAgent import name [as 别名]
# Construct an experiment to test the market.
experiment = MarketExperiment([], [], market)
# Add the agents and their tasks.
for g in case.generators:
env = DiscreteMarketEnvironment([g], market, dimState, markups, numOffbids)
task = ProfitTask(env)
module = ActionValueTable(dimState, dimAction)
module.initialize(1.0)
# learner = SARSA(gamma=0.9)
learner = Q()
# learner = QLambda()
# learner.explorer = BoltzmannExplorer() # default is e-greedy.
agent = LearningAgent(module, learner)
agent.name = g.name
experiment.tasks.append(task)
experiment.agents.append(agent)
# Prepare for plotting.
pylab.figure(1)#figsize=(16,8))
pylab.ion()
pl = MultilinePlotter(autoscale=1.1, xlim=[0, 24], ylim=[0, 1],
maxLines=len(experiment.agents))
pl.setLineStyle(linewidth=2)
pl.setLegend([a.name for a in experiment.agents], loc='upper left')
pylab.figure(2)
pylab.ion()
pl2 = MultilinePlotter(autoscale=1.1, xlim=[0, 24], ylim=[0, 1],
maxLines=len(experiment.agents))