Question

我想使用odeint将方波作为输入集成到耦合微分方程中。下面的代码似乎工作，但它会抛出一个错误：第12行的“IndexError：index out of bounds”。

任何人都可以帮我吗？

提前致谢，

Jyotika

   1 import pylab as pl
   2 import numpy as np
   3 from scipy.integrate import odeint
   4 
   5 
   6 def func1(Rates,time,ip):
   7         # 0 = x, 1 = y
   8         # dx/dt = -x+ ax+by, dy/dt =-y+ cx+dy
   9         leak = -0.3
   10         a = d = 0.1     
   11         b = c = -0.2    
   12         dxbydt = leak*Rates[0]+ a*Rates[0]+b/2.*Rates[1]+ip[time]
   13         dybydt = leak*Rates[1]+ c*Rates[0]+d*Rates[1]+ip[time]
   14                                 
   15         return [dxbydt, dybydt]

   16 time = np.arange(0,1000,1)
   17 ip = np.zeros((len(time)))
   18         
   19 ip[300:600] = 5.0
   20 initR = np.ones((2))*10
   21 fR = odeint(func1,initR,time,args=(ip,))
   22         
   23 pl.figure()
   24 #pl.plot(time,ip,'k-',label='ip')
   25 pl.plot(time,fR[:,0],'b-',label='x')
   26 pl.plot(time,fR[:,1],'r-',label='y')
   27 pl.legend()
   28 
   29 pl.show()

Answer 1

问题的关键是ip[time]，因为time的值在函数内部不断变化，这是重复调用的。 time增加，然后达到ip的最大索引值，即1000。会出现问题，因为时间是浮点数，但它被用作ip的索引。这里time的目的应该是什么？

Answer 2

import numpy as np
from scipy.integrate import odeint

def getIp(time):
        if time > 300 and time < 600:
                x = 5
        else:
                x = 0
        return x
def func1(Rates,time):
        # 0 = x, 1 = y
        # dx/dt = -x+ ax+by, dy/dt =-y+ cx+dy
        leak = -0.3
        a = d = 0.1
        b = c = -0.2
        dxbydt = leak*Rates[0]+ a*Rates[0]+b/2.*Rates[1]+getIp(time)
        dybydt = leak*Rates[1]+ c*Rates[0]+d*Rates[1]+getIp(time)
        return [dxbydt, dybydt]

timeGrid = np.arange(0,1000,0.01)
ip = np.zeros((len(timeGrid)))

#ip[300:600] = 5.0
initR = np.ones((2))*10
fR = odeint(func1,initR,timeGrid)

pl.figure()
pl.plot(timeGrid,ip,'k-',label='ip')
pl.plot(timeGrid,fR[:,0],'b-',label='x')
pl.plot(timeGrid,fR[:,1],'r-',label='y')
pl.legend()
pl.show()

scipy.integrate.odeint中的时间相关输入？

2 个答案: