How to get pairwise distance for two lists of vectors in python

Asked Dec 28 '20 at 15:31

Active Dec 28 '20 at 15:31

Viewed 180 times

Say I got two lists of vectors:

l1 = [v1, v2, v3]
l2 = [v4, v5, v6]

I know that with scipy I can get pairwise distances within a list but how do I do it for two so that the results look like that:

[[d(v1, v4), d(v1, v5), d(v1,v6)],
 [d(v2, v4), d(v2, v5), d(v2, v6)],
 [d(v3, v4), d(v3, v4), d(v3, v4)]]

in the most efficient way possible? loops are best to be avoided.

asked Dec 28 '20 at 15:31

Mario

3

Check [cdist](https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.distance.cdist.html) – Dani Mesejo Dec 28 '20 at 15:34
@DaniMesejo yes I know about cdist and pdist, Im asking how to use them for this particular arrangement – Mario Dec 28 '20 at 15:45
Well cdist works for two datasets, you can convert the list to numpy arrays and use cdist on them. – Dani Mesejo Dec 28 '20 at 15:47
i will look into that, thanks – Mario Dec 28 '20 at 15:48
And...? Got it going for you? – Willem Hendriks Dec 29 '20 at 21:02

0 Answers0