[pymvpa] getting simple regression working

[kimberg at mail.med.upenn.edu]

Emanuele replied:
> GPR with linear kernel is linear regression. It is Bayesian linear 
> regression as shown here (chapter 2.1):...

Thanks for the link, looks like that book might be well written enough for me to understand.  I admit that my understanding of GPR is superficial at best, but it looks like I had it misconfigured, with sigma_noise at its default of 0.001.  When I set it to a smaller value, I get residuals comparable to what I see with R, around 1e-12.

> Anyway they are in my 2010 todo/whish list and you are warmly invited
> to participate! :-)

I haven't gotten too deeply into the code yet, but I'll certainly keep that in mind as I start to work on some real problems.

Thanks again,

dan