time-series analysis in Gnumeric

[Frederic Parrenin <parrenin ujf-grenoble fr>]

Hi all,

With Franck Laurency, we are trying to add some time-series analysis
fonctionnalities to gnumeric, e.g.:
- a tool to resample (x_i,y_i) series at (x'_i) values, doing simple
interpolation or averaging, and the interpolant function being
staircase, linear, or cubic spline.
- Various frequency analysis tools (MTM, B-Tukey, Periodogram, Max.
Entropy)
- Various filtering tools
- Various smoothing tools
- a correlation tool (to analyse phases leads and lags)
- a tool to compute principal components
- etc.
The idea is to add some sheet functions, and then to make simple and
intuitive GUIs to these functions. 
This is not meant to be exhaustive time-series analysis
fonctionnalities, but just basic fonctionnalities for people who don't
want to learn more complex softwares like R or Matlab and that are
familiar with XL. 

We have already started a resampling tool and we have a prototype:
http://www-lgge.obs.ujf-grenoble.fr/~parrenin/download/gnumeric-tseries/screenshot1.png
http://www-lgge.obs.ujf-grenoble.fr/~parrenin/download/gnumeric-tseries/screenshot2.png
http://www-lgge.obs.ujf-grenoble.fr/~parrenin/download/gnumeric-tseries/screenshot3.png
http://www-lgge.obs.ujf-grenoble.fr/~parrenin/download/gnumeric-tseries/screenshot4.png

For now, the GUI tools have been added to the statistical tools, because
it is the same kind of GUI that we want.
Now, we want to move this code in a plugin.
(By the way, why the statistical tools are not coded as a plugin?)

Do you think it is more appropriate to write the GUI part in python or
in C?
For example, how complete is the python interface for plugins?
Also, is it easy to reuse the code we used in the statistical tools?

Is there any documentation on how to do that (either in python or in C)?

Cheers,

Frédéric PARRENIN and Franck LAURENCY