How can I fix “IndexError: only integers” in a numpy array?

(Solved. See answer below) I am teaching myself Python in preparation for a school project. I ran into the following error that I cannot figure out how to fix:

IndexError: only integers, slices (:), ellipsis (...), numpy.newaxis (None) and integer or boolean arrays are valid indices

:

...

None

Here is the code I am trying to run (edited it some based on comments)

from sympy import *
from sympy.abc import i
from mpmath import *
import numpy as np
np.set_printoptions(suppress=True)

N = 3
M = np.zeros(shape=(N, N))
for n in range(0, N):
M[0, n] = fac(n) # mpmath.fac
for m in range(1, N):
for n in range(0, N):
M[m, n] = (Sum(M[m - 1, i], (i, 0, n + 1))) / (n+1)
print(M)


Essentially, I want to specify a dimension for the array and a first row for the array. Then I want construct the remaining entries in the array in a certain way (Holder summability if you know about that summation technique).

I have looked over various other questions in stack overflow with this particular error but the questions were beyond my current knowledge or did not exactly address my issue. I have fiddled with many of the parameters and have isolated the issue to line 13. In particular, as soon as I put "i" inside M[m-1,i], I receive the error. My code runs if I put any number less than the dimension instead of "i" but this is not the construction I need.

For example, the entry M[2,1] should be constructed as:

M[2,1]

M[2,1] = (M[1,0] + M[1,1]) / 2

M[2,1] = (M[1,0] + M[1,1]) / 2

Which I thought corresponded to the sum I had written. How did my indexes become something other than a positive integer?

Thanks!

M[m - 1, i]

sympy.abc

i

i

Sum

M

M[m - 1, i]

2 Answers
2

You are mixing numpy and sympy, and it's giving problems:

numpy

sympy

Your sympy import defines i as a symbol (that's not obvious to numpy users):

sympy

i

numpy

In [1]: from sympy.abc import *
In [2]: i
Out[2]: i


then you define a numpy array. (import numpy as np is better than *):

numpy

import numpy as np

*

In [3]: M=np.zeros((3,3))


trying to use this i as index raises your error:

i

In [4]: M[1,i]
------------------------------------------------------------------
IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices


A symbol may work in a sympy matrix, but it does not work when indexing a numpy array.

sympy

Ms = sy.zeros(3,3) # sympy matrix
for m in range(1,3):
for n in range(3):
Ms[m,n] = (sy.Sum(Ms[m-1,i],(i,0,n+1)))/(n+1)


This runs, though the resulting Ms doesn't make much sense.

Ms

So the root of the problem is that you are mixing numpy and sympy, possibly through ignorance, and possibly because of the * imports.

*

I was able to fix the error based on everyone's help here. Thanks! I see that I should not star-import everything and that I should stick with either sympy or numpy so they aren't fighting each other. I was successfully able to rewrite my program to correctly build the Holder sums. Here is my running code:

import sympy as sy
import mpmath as mp
import numpy as np
np.set_printoptions(suppress=True)
sy.init_printing(use_unicode=False, wrap_line=False, no_global=True)
N = 5
M = np.zeros((N, N))
for n in range(0, N):
M[0, n] = mp.fac(n)
for n in range(0, N):
for i in range(0, n + 1):
M[1, n] += (M[0, i])
for m in range(2, N):
for n in range(0, N):
for i in range(0, n + 1):
M[m, n] += (M[m - 1, i])
M[m, n] = M[m, n] / (n + 1)
print(M)


This correctly (and beautifully) prints:

[[ 1. 1. 2. 6. 24. ]
[ 1. 2. 4. 10. 34. ]
[ 1. 1.5 2.33333333 4.25 10.2 ]
[ 1. 1.25 1.61111111 2.27083333 3.85666667]
[ 1. 1.125 1.28703704 1.53298611 1.99772222]]


Thanks again for the help!

Thanks for contributing an answer to Stack Overflow!

But avoid …

To learn more, see our tips on writing great answers.

Required, but never shown

Required, but never shown

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.