Suppose I've binned some data in a structure like this:
data = {(1,1): [...] # list of float,
(1,2): [...],
(1,3): [...],
(2,1): [...],
... }
here I've only two axis for the binning, but suppose I've N of them. Now suppose for example I have N=3 axis and I want the data where the second bin is 1, so I want a function
(None, 1, None) -> [(1, 1, 1), (1, 1, 2), (1, 1, 3), ...
(2, 1, 1), (2, 1, 2), (2, 1, 3), ...]
so I can use itertools.chain
for the result
you know the range of every axis from:
axes_ranges = [(1, 10), (1, 8), (1, 3)]
other examples:
(None, 1, 2) -> [(1, 1, 2), (2, 1, 2), (3, 1, 2), ...]
(None, None, None) -> all the combinations
(1,2,3) -> [(1,2,3)]
Seems very much like you reinvent the wheel. What you probably want to use is numpy.ndarray:
import numpy as np
>>> x = np.arange(0,27)
>>> x
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, 24, 25, 26])
>>> x.reshape(3,3,3)
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],
[24, 25, 26]]])
>>> x[0]
array([[0, 1, 2],
[3, 4, 5],
[6, 7, 8]])
>>> x[:,1,:]
array([[ 3, 4, 5],
[12, 13, 14],
[21, 22, 23]])
>>> x[:,1,1]
array([ 4, 13, 22])
This can have N dimensions. In the example the indexing is threedimensional, you can see it as a cube with x[a,b,c] = x[layer,row,column]. Using a ":" as index simply means "all"