[SciPy-user] kernel density estimation in scipy?

Pierre GM pgmdevlist@gmail....
Tue Aug 28 07:31:51 CDT 2007


On Tuesday 28 August 2007 05:50:17 massimo sandal wrote:
> The software works nicely, but it doesn't seem very well maintained (it
> says it has not been checked with NumPy, for example). Does a similar
> functionality exist on SciPy? If not, why not considering including it
> in scipy? Are there other KDE and related statistical packages for Python?

If you're willing to be a beta-tester, I'm working on adapting SiZer maps to 
Python.
http://www.stat.unc.edu/faculty/marron/DataAnalyses/SiZer_Intro.html
The module presents some functions to compute Gaussian KDEs, find the optimal 
bandwidth according to different algorithms (Terrell, Sheather-Jones, 
Rupert-Sheather-Wand). A couple of features are still missing (such as 
handling missing/censored values), but I should be able to send you the 
sources nevertheless. In the next few days (couple of weeks) I hope to be 
able to propose it as a Scikit


More information about the SciPy-user mailing list