On 13 Mar 2014 02:24, "Stuart Mumford" <stuart@mumford.me.uk> wrote:

I agree, I think it could be a combined scipy effort. Going to need some

Astropy magic for the data io (probably) and the Ra/Dec coordinate handling. Then we can use sklearn and potentially astroml once we have a detection routine? That sounds about right. Do you have ties to the AstroPy community? Perhaps someone there would be willing to lead the charge. Regards Stéfan