本文整理汇总了Python中Agent.Agent.decide方法的典型用法代码示例。如果您正苦于以下问题：Python Agent.decide方法的具体用法？Python Agent.decide怎么用？Python Agent.decide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Agent.Agent的用法示例。
在下文中一共展示了Agent.decide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: NormalARD
# 需要导入模块: from Agent import Agent [as 别名]
# 或者: from Agent.Agent import decide [as 别名]
lik = np.log(np.array([0.00001]))
hyp = np.log(np.array([1, 1, 10]))
cov = NormalARD()
gp = GaussianProcess(lik, hyp, cov)
gp2 = GaussianProcess(lik, hyp, cov)
sig =np.ones((3,)) * 0.001
sig2 = np.ones((3,)) * 0.1
start_z = np.array([[0., 0., 0.]])
agent = Agent(gp, reward, sig, start_z)
agent2 = Agent(gp2, reward, sig2, start_z)
fig = plt.figure(figsize=(20,7), dpi=300)
zlim = (-10, 10, -10, 10)
for i in xrange(0, 1000):
agent.observe()
agent.decide()
agent.act()
agent2.observe()
agent2.decide()
agent2.act()
t = agent.gp.Z[-1].flatten()[-1]
a = [0] * 4
a[0] = agent.gp.Z[-1].flatten()[0]
a[1] = agent.gp.Z[-1].flatten()[1]
a[2] = agent.gp.Z[-1].flatten()[0]
a[3] = agent.gp.Z[-1].flatten()[1]
extent = np.max(np.abs(a))
lim = extent + 3 if extent > 10 else 10
zlim = (-lim, lim, -lim, lim)
fig.clf()示例2: execute
# 需要导入模块: from Agent import Agent [as 别名]
# 或者: from Agent.Agent import decide [as 别名]
def execute(self):
##
## Initialize agents
##
pDisease = {Constant.BETA: 1 - math.exp(-self.disease[Constant.BETA]),
Constant.RHO: self.disease[Constant.RHO],
Constant.GAMMA: 1 - math.exp(-self.disease[Constant.GAMMA])}
self.decision = 1 - math.exp(-self.decision)
N = 0
agents = []
infected = []
for state in self.nAgents:
for x in range(self.nAgents[state]):
agent = Agent(N, state, pDisease, self.fear, self.timeHorizon, self.payoffs)
agents.append(agent)
if (state == State.I):
infected.append(agent)
N += 1
##
## Output variables
##
num = []
num.append([0,
self.nAgents[State.S],
self.nAgents[State.P],
0,
self.nAgents[State.I],
0,
0,
self.nAgents[State.R],
0,
0,
self.nAgents[State.S] * self.payoffs[State.S],
self.nAgents[State.P] * self.payoffs[State.P],
self.nAgents[State.I] * self.payoffs[State.I],
self.nAgents[State.R] * self.payoffs[State.R]])
##
## Run the simulation
##
t = 1
i = self.nAgents[State.I] / float(N)
while ((t < self.timeSteps) and (i > 0)):
numagents = [0, 0, 0, 0]
##
## Interaction
##
shuffle(agents)
n = N
infected = []
while(n > 1):
a1 = agents[n - 1]
a2 = agents[n - 2]
a1State = a1.getState()
a2State = a2.getState()
a1S = a1State
a2S = a2State
if (a1State == State.I):
infected.append(a1)
a2S = a2.interact(a1State)
if (a2State == State.I):
infected.append(a2)
a1S = a1.interact(a2State)
numagents[a1S] += 1
numagents[a2S] += 1
n = n - 2
##
## Decision
##
for agent in agents:
if (uniform(0.0, 1.0) < self.decision):
state = agent.getState()
numagents[state] -= 1
state = agent.decide(i)
numagents[state] += 1
##
## Recover
##
for agent in infected:
if (agent.recover() == State.R):
numagents[State.I] -= 1
numagents[State.R] += 1
#.........这里部分代码省略.........