[Numpy-discussion] argsort speed

[Charles R Harris]

On Sun, Feb 16, 2014 at 4:12 PM, Daπid <davidmenhur at gmail.com> wrote:

> On 16 February 2014 23:43, <josef.pktd at gmail.com> wrote:
>
>> What's the fastest argsort for a 1d array with around 28 Million
>> elements, roughly uniformly distributed, random order?
>>
>
> On numpy latest version:
>
> for kind in ['quicksort', 'mergesort', 'heapsort']:
>     print kind
>     %timeit np.sort(data, kind=kind)
>     %timeit np.argsort(data, kind=kind)
>
>
> quicksort
> 1 loops, best of 3: 3.55 s per loop
> 1 loops, best of 3: 10.3 s per loop
> mergesort
> 1 loops, best of 3: 4.84 s per loop
> 1 loops, best of 3: 9.49 s per loop
> heapsort
> 1 loops, best of 3: 12.1 s per loop
> 1 loops, best of 3: 39.3 s per loop
>
>
Interesting. I get slightly different results

quicksort
1 loops, best of 3: 3.25 s per loop
1 loops, best of 3: 6.16 s per loop
mergesort
1 loops, best of 3: 3.99 s per loop
1 loops, best of 3: 6.97 s per loop
heapsort
1 loops, best of 3: 10.1 s per loop
1 loops, best of 3: 29.3 s per loop

Possibly faster memory here.

<snip>

Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20140216/d6c8b1e8/attachment.html>