[Numpy-discussion] sort descending with NaNs

[Charles R Harris]

On Thu, Jan 27, 2011 at 10:10 AM, Charles R Harris <
charlesr.harris at gmail.com> wrote:

>
>
> On Thu, Jan 27, 2011 at 9:50 AM, Fabrizio Pollastri <f.pollastri at inrim.it>wrote:
>
>> Hello,
>>
>> when one has to find a given number of highest values in an array
>> containing
>> NaNs, the sort function (always ascending) is uncomfortable.
>>
>> Since numpy >= 1.4.0 NaNs are sorted to the end, so the searched values
>> are just
>> before the first NaN in a unpredictable position and one has to do another
>> search for the first NaN position.
>>
>> Sorting descending will solve the problem, but there is no option with
>> numpy
>> sort. There is any other trick to avoid this second search?
>>
>
> If you just want to reverse the result, try a[::-1]. I think you may still
> need to find the boundaries of the nan's just to make sure they aren't
> included among the largest values. Searchsorted is pretty quick in any case.
>
>
To sort in descending order sort the negatives, i.e.

In [1]: -sort(-array((0,1,2,3,4,nan)))
Out[1]: array([  4.,   3.,   2.,   1.,   0.,  nan])

I still think a.searchsorted(nan) would be faster.

Chuck
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/numpy-discussion/attachments/20110127/219f7026/attachment.html>