6

I am trying to use Kernel Density Estimation (KDE) to compute the pdf of sample data points of d-dimension. I have read the wiki page in which they cite library libAGF. However this site has no examples nor tutorials. I am reluctant to write the code myself as I'm sure a library would be more optimised (possibly using nearest neighbour search to speed up evaluation).

Does anyone know of any library which implements this functionality?

Charles
  • 50,943
  • 13
  • 104
  • 142
Aly
  • 15,865
  • 47
  • 119
  • 191
  • have you looked at the [libAGF's Quickstart file](http://libagf.cvs.sourceforge.net/viewvc/libagf/agf/QUICKSTART?revision=1.3&view=markup) ? – Andre Holzner Sep 17 '13 at 11:29
  • 6
    Why was this marked off topic? Anyway, when I needed it I just wrote it myself, complete with skipping the calculation for more distant neighbors. I could post the code, but.... – Michael Dec 20 '13 at 02:00
  • 1
    @Michael They close certain topics when posted for C++ that they don't close when posted for other languages. They have pretty horrible reasoning for it, which you can reference in my attempts at C++ encryption questions. Anyhow, if I get you contact information, could you send me your code? – Chris Redford Mar 30 '16 at 21:36
  • I have excel formulas... does that help? – Michael Mar 31 '16 at 19:39
  • @Michael Yes. I've started a chat for the topic [here](http://chat.stackoverflow.com/rooms/107975/kde-estimation-chat). – Chris Redford Apr 01 '16 at 17:05

0 Answers0