Raymond Hettinger <raymond.hettin...@gmail.com> added the comment:
> Are you happy guaranteeing that it will always be the first
> mode encountered?
Yes.
All of the other implementations I looked at make some guarantee about which
mode is returned. Maple, Matlab, and Excel all return the first encountered.¹
That is convenient for us because it is what Counter(data).most_common(1)
already does and does cheaply (single pass, no auxiliary memory). It also
matches what a number of our other tools do:
>>> max(3, 3.0) # 3 is first encountered
3
>>> max(3.0, 3) # 3.0 is first encountered
3.0
>>> list(dict.fromkeys('aabbaacc'))[0] # 'a' is first encountered
'a'
>>> sorted([3, 3.0])[0] # 3 is first encountered (due to sort stability)
3
>>> sorted([3.0, 3])[0] # 3.0 is first encountered (due to sort stability)
3.0
¹ Scipy returned the smallest value rather than first value but it algorithm
was sorting based to accommodate running a parallel mode() computation on
multiple columns of an array. For us, that approach would be much slow, would
require more memory, and would require more bookkeeping.
P.S. I'm no longer thinking this should be saved for Pycon sprints. That is
right at the beta 1 feature freeze. We should aim to get this resolved well in
advance of that.
----------
_______________________________________
Python tracker <rep...@bugs.python.org>
<https://bugs.python.org/issue35892>
_______________________________________
_______________________________________________
Python-bugs-list mailing list
Unsubscribe:
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
