Calculate mean of each 2d array in a numpy array

Question

I have an numpy array as following:

b = numpy.array([[[1,2,3], [4,5,6]], [[1,1,1],[3,3,3]]])
print(b)
[[[1 2 3]
  [4 5 6]]

 [[1 1 1]
  [3 3 3]]]

now I wan't to calculate the mean of each 2-d array in the array. for e.g.

numpy.mean(b[0])
>>> 3.5
numpy.mean(b[1])
>>> 2.0

How do I do this without using a for loop?

`np.mean(b, axis=0)` or simply `b.mean(axis=0)` or `axis=1` depending if you want to average rows or columns. — Itay, Aug 20 '20 at 06:17
Does this answer your question? [Calculate mean across dimension in a 2D array](https://stackoverflow.com/questions/15819980/calculate-mean-across-dimension-in-a-2d-array) — RichieV, Aug 20 '20 at 06:25
No, none of these address the question I had. Please note that I have an array of 2D arrays. Not a 2D array. — Saqib Ali, Aug 20 '20 at 06:37
https://numpy.org/doc/stable/reference/generated/numpy.mean.html — Paul H, Aug 20 '20 at 06:41

Dishin H Goyani · Accepted Answer · 2020-08-20T06:42:18.657

4

I think this would give you your expected output

b.mean(axis=(1,2))
array([3.5, 2. ])

edited Aug 20 '20 at 06:42

answered Aug 20 '20 at 06:31

Dishin H Goyani

skip the loop. no point in looping for an operational like this on a numpy array – Paul H Aug 20 '20 at 06:35
Thanks. This is exactly what I was looking for. I understand the FOR loop, but can you please explain the first solution you described? – Saqib Ali Aug 20 '20 at 06:36
@PaulH yes, removed it. – Dishin H Goyani Aug 20 '20 at 06:38
1

@SaqibAli It will performed mean on multiple dims as passed. – Dishin H Goyani Aug 20 '20 at 06:40

score 2 · Answer 2 · answered Aug 20 '20 at 06:28

2

np.mean() can take in parameters for the axis, so according to your use, you can do either of the following

print("Mean of each column:")
print(x.mean(axis=0))
print("Mean of each row:")
print(x.mean(axis=1))

answered Aug 20 '20 at 06:28

Ravish Jha

To expand on this: Axis of 0 is rows, 1 is columns. You iterate over rows to compute the sum of each column and you iterate over columns to compute the sum of each row which is how you get the code above. – J.Doe Sep 05 '21 at 03:06

2 Answers2