Repository navigation
random choice should support multidimensional arrays #2724
Description
random.choice enhancement request:
random.choice should support multidimensional arrays by taking an axis number. With this enhancement, axis=None would choose from a flattened array, while an integer argument would chose from the subarrays along that axis.
Apologies if this is a duplicate. My previous submission seems to have vanished.
Activity
seberg commentedon Nov 12, 2012
This should be simple if you are thinking about doing it. All that seems necessary is to add a special check for
axis=Noneand ravel in that case and replace makeaxis=1. Then setpop_size = a.shape[axis]and passaxisintotake. (Plus possibly an error ifais just anintand axis is given).Actually thinking about it, it could also be reasonable to have
size=Noneas default and have it return a scalar/the given axes removed in that case. That would be a real change in functionality unfortunately, but it would make sense for similarity toimport random; random.choicemostly.One could think about also allowing an
outargument which is just passed intotakemostly. While none of the other random functions have that it could be a nice performance option in some cases.ENH: Add multidimensional array support to numpy.random.choice
samuela commentedon Nov 23, 2016
+1
jhollowed commentedon Feb 28, 2017
+1. Is this still not implemented?
jbachh commentedon Mar 5, 2018
+1
slidenerd commentedon Apr 22, 2018
Is this implemented?
mikesmith1611 commentedon Apr 30, 2018
Is this ever going to be implemented?
JulianKlug commentedon Jun 19, 2018
+1
ENH: Add axis functionality to random.choice
stephenpardy commentedon Jul 11, 2018
+1
I also tried implementing this, before realizing that @masasin had already done it.
mattip commentedon May 29, 2019
Implemented in the new randomgen API:
np.random.Generator().choice(..., axis=int). Closing since we will not be updating the legacy code to support this. If you feel differently, please reopen for discussion.