URL(s) with the issue:
Description of issue (what needs changing):
The documentation for the cosine similarity does not state whether
As a special case, the doc does not state what happens in the case of either being zero.
(Also, isn't the above implementation suboptimal in terms of speed, as each element is divided by
Yes, can you send a PR?
The code is here:
If you make a PR please use doctest format, for example code so that it is testable (prefix code lines with