efficient way to process data

[Petite Abeille]

On Jan 13, 2014, at 7:42 PM, Mark Lawrence <breamoreboy at yahoo.co.uk> wrote:

> I've not followed this thread closely but would this help http://pandas.pydata.org/ ?  When and if you get back to it, that is!!!

I doubt it. The mean overhead by far would be to shuffle pointless data between the server & client. Best to process data closest to their source.

On the other hand:

http://devour.com/video/never-say-no-to-panda/