在预先分配单元阵列方面遇到很多麻烦'在Python中

时间:2015-11-04 21:36:07

标签: python arrays matlab numpy

我最近从MATLAB的编程转向Python编程。因此,我遇到了一些运行我编写的Python代码的问题。我使用numPy和SciPy来集成任意一组常微分方程,方法 vode 。我已经概括了任何数字的ODE系统&n;其中,但遇到预分配和使用数组的问题。

对我来说尤其令人沮丧,因为我有两个版本的全功能MATLAB代码,但需要将其转换为Python以获得最佳结果。我遇到了麻烦。特别是以下几行:

S = np.array(np.zeros((N/2+1,1)), dtype = 'object')
KS = np.array(np.zeros((N/2+1,1)), dtype = 'object')
PS = np.array(np.zeros((N/2+1,1)), dtype = 'object')
Splot = np.array(np.zeros((N/2+1,1)), dtype = 'object')
KSplot = np.array(np.zeros((N/2+1,1)), dtype = 'object')
PSplot = np.array(np.zeros((N/2+1,1)), dtype = 'object')

以下是代码:

import numpy as np
import matplotlib.pyplot as plt
from scipy import integrate

N = 10
K00 = np.logspace(0,3,101,10)
len1 = len(K00)
epsilon = 0.01
y0 = [0]*(3*N/2+3)
u1 = 0
u2 = 0
u3 = 0
Kplot = np.zeros((len1,1))
Pplot = np.zeros((len1,1))
S = np.array(np.zeros((N/2+1,1)), dtype = 'object')
KS = np.array(np.zeros((N/2+1,1)), dtype = 'object')
PS = np.array(np.zeros((N/2+1,1)), dtype = 'object')
Splot = np.array(np.zeros((N/2+1,1)), dtype = 'object')
KSplot = np.array(np.zeros((N/2+1,1)), dtype = 'object')
PSplot = np.array(np.zeros((N/2+1,1)), dtype = 'object')

for alpha in range(0,(N/2+1)):
    Splot[alpha] = np.zeros((len1,1))
for beta in range((N/2)+1,N+1):
    KSplot[beta-N/2-1] = np.zeros((len1,1))
for gamma in range(N+1,3*N/2+1):
    PSplot[gamma-N] = np.zeros((len1,1))

for series in range(0,len1):
    K0 = K00[series]
    Q = 10
    r1 = 0.0001
    r2 = 0.001
    a = 0.001
    d = 0.001
    k = 0.999
    S10 = 1e5
    P0 = 1

    def f(y, t):
        for alpha in range(0,(N/2+1)):
            S[alpha] = y[alpha]
        for beta in range((N/2)+1,N+1):
            KS[beta-N/2-1] = y[beta]
        for gamma in range(N+1,3*N/2+1):
            PS[gamma-N] = y[gamma]
        K = y[3*N/2+1]
        P = y[3*N/2+2]

        ydot = np.zeros((3*N/2+3,1))
        B = range((N/2)+1,N+1)
        G = range(N+1,3*N/2+1)
        runsumPS = 0
        runsum1 = 0
        runsumKS = 0 
        runsum2 = 0

        for m in range(0,N/2):
            runsumPS = runsumPS + PS[m+1]
            runsum1 = runsum1 + S[m+1]
            runsumKS = runsumKS + KS[m]
            runsum2 = runsum2 + S[m]    
            ydot[B[m]] = a*K*S[m]-(d+k+r1)*KS[m]

        for i in range(0,N/2-1):
            ydot[G[i]] = a*P*S[i+1]-(d+k+r1)*PS[i+1]

        for p in range(1,N/2):
            ydot[p] = -S[p]*(r1+a*K+a*P)+k*KS[p-1]+ \
                      d*(PS[p]+KS[p])

        ydot[0] = Q-(r1+a*K)*S[0]+d*KS[0]+k*runsumPS
        ydot[N/2] = k*KS[N/2-1]-(r2+a*P)*S[N/2]+ \
                    d*PS[N/2]
        ydot[G[N/2-1]] = a*P*S[N/2]-(d+k+r2)*PS[N/2]
        ydot[3*N/2+1] = (d+k+r1)*runsumKS-a*K*runsum2
        ydot[3*N/2+2] = (d+k+r1)*(runsumPS-PS[N/2])- \
                        a*P*runsum1+(d+k+r2)*PS[N/2]

        for j in range(0,3*N/2+3):
            return ydot[j] 

    if __name__ == '__main__':

        r = integrate.ode(f).set_integrator('vode', method='bdf')  
        t_start = 0.0
        t_final = 1e10
        delta_t = t_final/(len1-1)
        num_steps = np.floor((t_final - t_start)/delta_t) + 1

        y0[0] = S10
        for i in range(1,3*N/2+1):
            y0[i] = 0
        y0[3*N/2+1] = K0
        y0[3*N/2+2] = P0
        r.set_initial_value(y0, t_start)

        t = np.zeros((num_steps, 1))
        soln = np.array(np.zeros((num_steps, 1))*(3*N/2+3))
        t[0] = t_start
        for i in range(0,3*N/2+3):
            soln[i] = y0[i]

        k = 1
        while r.successful() and k < num_steps:
            r.integrate(r.t + delta_t)

            t[k] = r.t
            for jj in range(0,3*N/2+3):
                soln[k] = r.y[jj]
            k += 1

错误信息如下:

ValueError                                Traceback (most recent call last)
C:\Users\dis_YO_boi\Documents\Programming\Python\ArrayMod.py in <module>()
     21 
     22 for alpha in range(0,(N/2+1)):
---> 23     Splot[alpha] = np.zeros((len1,1))
     24 for beta in range((N/2)+1,N+1):
     25     KSplot[beta-N/2-1] = np.zeros((len1,1))

ValueError: could not broadcast input array from shape (101,1) into shape (1)

提前感谢您的帮助。

1 个答案:

答案 0 :(得分:3)

在尝试完全理解您的代码之前,让我观察一下这样的结构:

S = np.array(np.zeros((N/2+1,1)), dtype = 'object')

不好numpy

您可能正在模仿MATLAB单元阵列。 Python在MATLAB之前很久就有'单元'。 Python列表可以包含多种值,字符串,数字,其他列表,数组等。

带有numpy

dtype=object数组只是美化列表。如果你想要2d集合,它们可能很方便,但与MATLAB一样,你不能对这些数组的元素进行数学运算。充其量你可以迭代它们,就像你对列表一样。

你的错误可能与此无关,但我必须要挖掘一下才能确定。

Splot=np.array(np.zeros((4,1)),dtype=object)

生成一个(4,1)数组,其中dtype = object。 np.array尝试从输入创建尽可能高的维数组。

从     对于范围内的alpha(0,(N / 2 + 1)):         Splot [=] = np.zeros((len1,1))

看起来您想要预先分配一个包含这些N/2+1个插槽的数组,并为每个插槽填充一个2d数组。这对dtype=object来说有点棘手。

Splot = [np.zeros((len1,1)) for alpha in range(M)]

会生成一个包含M个数组的列表,每个数组的长度相同。

e.g。

In [67]: Splot=[np.zeros((4,1)) for alpha in range(3)]

In [68]: Splot
Out[68]: 
[array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]]),
 array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]]),
 array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]])]

请注意,我将这个2d数组列表包装在一个数组中,我得到一个3d数组:

In [69]: np.array(Splot)
Out[69]: 
array([[[ 0.],
        [ 0.],
        [ 0.],
        [ 0.]],

       [[ 0.],
        [ 0.],
        [ 0.],
        [ 0.]],

       [[ 0.],
        [ 0.],
        [ 0.],
        [ 0.]]])

In [70]: _.shape
Out[70]: (3, 4, 1)

可以使用此策略制作数组数组。

创建一个大小合适的对象数组(np.empty将用None填充它)。

In [72]: Splot
Out[72]: array([0, 0, 0], dtype=object)

然后迭代以用新对象替换每个0

In [73]: for i in range(3):
   ....:     Splot[i] = np.zeros((4,1))
   ....:     

In [74]: Splot
Out[74]: 
array([array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]]),
       array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]]),
       array([[ 0.],
       [ 0.],
       [ 0.],
       [ 0.]])], dtype=object)