Boost logo

Boost :

From: Hans Larsen (hans_at_[hidden])
Date: 2007-03-16 17:38:20


You're right. I was considering nth_element behavior exactly the
same as partial_sort.

My bad. :)
Hans

On 16-Mar-07, at 5:11 PM, Michael Fawcett wrote:

> On 3/16/07, Hans Larsen <hans_at_[hidden]> wrote:
>>
>> On 16-Mar-07, at 4:42 PM, Michael Fawcett wrote:
>>
>>> On 3/16/07, Hans Larsen <hans_at_[hidden]> wrote:
>>>> Maybe finding the nth element of an array, without actually sorting
>>>> the array. This could be useful for finding 2nd min/max and median
>>>> of an array reusing the same algorithm.
>>>>
>>>> This is actually quite easy, and may have already been
>>>> implemented in
>>>> boost or stl (but i haven't found it yet).
>>>
>>> std::nth_element perhaps?
>>
>> Not exactly. nth_element merely takes the pivot that you give and
>> put it into place.
>>
>> What I'm looking at is a function that, given a position, returns the
>> element that correspond to that position when array is sorted.
>>
>> Hans
>>
>> Exemple:
>> template< class V >
>> V::value_type& get_median( V& vec )
>> {
>> return boost::sorting::__func__( vec.begin(), vec.end(),
>> vec.size
>> () / 2 );
>> }
>
> Sorry, I guess I'm having trouble understanding how that is different.
> From MSDN article on nth_element:
>
> "Partitions a range of elements, correctly locating the nth element of
> the sequence in the range so that all the elements in front of it are
> less than or equal to it and all the elements that follow it in the
> sequence are greater than or equal to it."
>
> In your example, nth_element would return the median value. In other
> words, the value that would be located at vec.size() / 2 if vec were
> completely sorted.
>
> --Michael Fawcett
> _______________________________________________
> Unsubscribe & other changes: http://lists.boost.org/mailman/
> listinfo.cgi/boost


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk