29 July 2011

R You Experienced?

Herewith some musings on politics and data.  Don't be afraid, the data doesn't bite.  Those interested in the relationship between the two could find this amusing.

I've always been conflicted:  data for analysis (stats) or data for use (RDBMS)????  I started adult life as an econometrician, and ended up doing the Relational Model.  Until recently, that was OK.  The problem is that the Lunatic Right gets ever more aggressive in its lying.  So, doing good with data is more important than the technology used, although smart database is better than dumb files.

Which brings me to R (a stat language).  I've been working with it for a couple of years, in an informal way.  Unlike SPSS, SAS, PSTAT, and the like, the syntax looks more like C or java than SQL.  Chapter 10 of "R in a Nutshell" is titled "Object-Oriented Programming".  So, a programmable stat language.  It also does graphics very cool.  It has bindings to most languages you'd find in the PC/server world (none for COBOL that I could find), an RODBC driver, an RJDBC driver (they do what you think), and some database specific drivers (Oracle and PostgreSQL, at least).

I had a chat with some folks, which chat was supposed to be about my data geekiness, but turned out to be about their code geekiness.  Which led to a conversation best described as an oxymoron; McElhone (a math stat I worked for some years ago) would have just said, "happily married".  It's kind of too bad; there's lots of data wonking to be done if 2012 isn't going to be worse than 2010.

Not having heard anything, despite emails saying I'd be kept in the loop, I've spent part of today looking for R topics on point, just to see what might have been.  Turns out, there's bunches.

First, there's R-bloggers, a blog aggregator.  The right hand column goes most of the way down the Yellow Brick Road.  Talk about bunches.  There's also PlanetR, but it's not so user friendly.

Second, there's this blog post I got to from R-bloggers.  This is the second post; the first (linked in the post) discussed the ruby mod which prepares the data.  The problem being solved is rather different (other than being about money, but that's the basis of any politics issue) from the one I attempted to chat about.  What is disturbing:  my interrogators evinced no knowledge of this.  It's more than 2 years ago, tied to ruby (their language of preference), and animated graphics (the sort of eye-candy they alluded to).

To quote:  "The entire process, from extraction to visualization was created in about half a day, so there is obviously more work that could be done."  That includes installing MySql and R; my interrogators didn't do MySql, but some other R compatible database, and didn't seem to know anything about R, although I had mentioned this when we chatted by phone days previously.  A half day.  Repeat, retch, wish one had voted for McCain.

Third, there continues to be, what I'll call, The Brazile Problem:  Democrats not knowing, or worse not speaking, the truth when the Lunatic Right spews self-serving shit.  Only the truth will make us free.

No comments: