That was my thought...but for some reasons some operations are very
slow from the fTools code, even if they use indexing. I have a couple
of examples with points in polygon calculation, that take ages in
fTools (or the same SEXTANTE algorithm, which has the same code), and
they shouldn't. Maybe the Python call to create a spatial index is not
working properly and it actually does nothing... ¿?¿?

I haven't spent too much time on this anyway, so it's true that
improving those algorithms might be possible and probably in some
cases not hard. I do not remember much about all ftools algorithms,
but for instance, those in the MMQGIS plugin were in many cases very
naive (and thus, unusable for large layers). I put them into SEXTANTE
as they were, but it's true that some extra work should be done on
that, at least with the most used algoirthms

Regards
Victor


2013/4/18 Martin Dobias <[email protected]>:
> On Thu, Apr 18, 2013 at 4:51 PM, Werner Macho <[email protected]> wrote:
>> hi!
>>
>> If I remember correct Carson Farmer startet to bring the algorithms into a
>> cpp library (which would be a good and fast choice i think)
>> I'd welcome having fast and reliable algorithms in cpp .. So I agree with
>> Agustin - very welcome work..
>
> Let me add that plain re-writing from Python to c++ may _not_ bring a
> great speed improvement. Often it is more important to change the
> algorithm - e.g. to use a spatial index where possible.
>
> Martin
> _______________________________________________
> Qgis-developer mailing list
> [email protected]
> http://lists.osgeo.org/mailman/listinfo/qgis-developer
_______________________________________________
Qgis-developer mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Reply via email to