Smoothing a discrete set of data

[Tim Churches]

Fernando Pérez wrote:
> 
> Tim Churches wrote:
> 
> > AFAIK, there is no native Python package which
> > implements a
> > comprehensive range of regression and smoothing techniques, but I can
> > highly
> > recommend the R package for statistics and statistical graphics - this
> > mature, free, open
> > source package has more statistical facilities than you are ever likely
> > to need.
> 
> While not as comprehensive as R in its statistical coverage, SciPy probably
> has enough in it for what the OP wants (which sounded fairly basic). Quick
> linear (in the parameters) fits are easily done with scipy, plotting, simple
> smoothing (if he wants to), etc. I know R is industrial-strength statistics,
> but it sounds like more than what the OP needed.

Yup, although I suggested R because a) some very gentle companion
introductions 
to basic statistical analysis are available for R (see for example
http://cran.r-project.org/doc/manuals/R-intro.pdf or the various
documents listed
at http://cran.r-project.org/other-docs.html ); and b) R will not run
out of steam
when the OP realises that there might be a wee bit more to the analysis
of his
data than he first thought eg some of his data are almost certainly time
series 
and hence he will need access to techniques which take auto-correlation
etc into
account. 

Of course, the biggest hurdle for Pythonistas who want to use R is the
fact that 
R uses one-based indexing, not zero-based indexing as <insert deity or
demi-god 
of your choice> intended.

Tim C