Background: Math-y with R/MATLAB experience; some work using Python/SQLAlchemy/Rails/d3 but I'm really an analyst, notsomuch a developer or DB guy (but always learning!). Looking at a DB that's <10GB, so not doing Hadoop/Hive-level stuff yet.
Haven't really tried my hand at the two together... typically I receive exports in csv files that I then play around with in R, but I think the better practice is mining the data directly without exporting, so I wanted to check what good practices/libraries people would recommend.
Would love to hear any advice or just points toward the right direction.
PS -- Bonus q, and a bit more ambitious -- would love to hear advice on dashboarding (whether it's d3-based, opensourced, or a freemium/enterprise solution) for such an SQL database as well. This is for a personal project so I'm preferring to bootstrap with time/labor. :)
Edit 4/3/13: Thanks guys for advice; you're all awesome!
[link][4 comments]