On 11/20/2011 12:57 AM, haskell heath wrote:
[ I've dabbled in lots of stuff, which language should I use for what?]
and
I'm new to Haskell and I can't really call myself a decent programmer in any other language. Do you think it's wrong to think that I can contribute to the statistics library?
I'm a journeyman R coder and a bare novice at Haskell. What springs to my mind here is, what's your primary goal: to enhance the statistical toolset available in Haskell, or to accomplish a task? If the former, then I think the critical question isn't your language competence but your statistical props; if you watch the R devel list for any duration, you'll see how deeply the real stats folks treat these problems; to have a broken tool (and not know it) is often worse than to have no tool. If you've got the stats clue, then by all means soldier on. Subject Matter Experts rock. :) If your goal is to accomplish your task, then I suggest that R is absolutely the superior environment for statistical thinking these days. Cobble together whatever data-collection bits you need in whatever toolset is convenient, and if you're scraping web, then node.js is as good a place to start as any. Drop CSV files from your scraping, and go to town in R. - Allen S. Rout