# resize with averaging or rebin a numpy 2d array

I am trying to reimplement in python an IDL function:

http://star.pst.qub.ac.uk/idl/REBIN.html

which downsizes by an integer factor a 2d array by averaging.

For example:

>>> a=np.arange(24).reshape((4,6)) >>> a array([[ 0, 1, 2, 3, 4, 5], [ 6, 7, 8, 9, 10, 11], [12, 13, 14, 15, 16, 17], [18, 19, 20, 21, 22, 23]])

I would like to resize it to (2,3) by taking the mean of the relevant samples, the expected output would be:

>>> b = rebin(a, (2, 3)) >>> b array([[ 3.5, 5.5, 7.5], [ 15.5, 17.5, 19.5]])

i.e. b[0,0] = np.mean(a[:2,:2]), b[0,1] = np.mean(a[:2,2:4]) and so on.

I believe I should reshape to a 4 dimensional array and then take the mean on the correct slice, but could not figure out the algorithm. Would you have any hint?

## Answers

Here's an example based on the answer you've linked (for clarity):

>>> import numpy as np >>> a = np.arange(24).reshape((4,6)) >>> a array([[ 0, 1, 2, 3, 4, 5], [ 6, 7, 8, 9, 10, 11], [12, 13, 14, 15, 16, 17], [18, 19, 20, 21, 22, 23]]) >>> a.reshape((2,a.shape[0]//2,3,-1)).mean(axis=3).mean(1) array([[ 3.5, 5.5, 7.5], [ 15.5, 17.5, 19.5]])

As a function:

def rebin(a, shape): sh = shape[0],a.shape[0]//shape[0],shape[1],a.shape[1]//shape[1] return a.reshape(sh).mean(-1).mean(1)