UPDATE: Last week, I got a fantastic book. It is, hands down, the best resource on linear modeling that I’ve come across. I’ve been devouring the pages and there’s still loads of stuff that I’ve not gotten round to yet. Suddenly, everything seems very, very clear. There’s even a nice refresher on linear algebra in the back. Great, great book. I’m using the second edition, which is described here: Introduction to the Theory and Practice of Econometrics. I picked up a barely used copy from a guy named Uncle Pinky. He’s got loads of great stuff and his Amazon site is worth checking out.
At the beginning of 2012 I started a new position at a fantastic company, which happens to have a very friendly and helpful IT department. I suddenly found myself with loads of freedom explore open source software, particularly R. This, in turn, ushered in a frenzy of book buying to learn more about statistical concepts and techniques that I’d never been able to explore with the tools I’d previously had (i.e. Excel). Here’s what I’ve been reading:
This is a fantastic book and I’ve barely been able to scratch the surface. Among other things, it’s an easy go-to reference for GLMs, a topic about which I’ve had no formal training whatsoever.
The title is a bit silly, but the content is wonderful. Piles of great R code and practical examples of interesting analytic techniques. Just an overview, but a very good one.
As anyone who knows me will tell you, I love SQL. That’s why I find the NoSQL movement so fascinating. This is a great way to understand the public gripes about SQLs limitations.
This is another by Andrew Gelman. It’s a great one to share with folks what aren’t familiar with statistical research. Particularly if they’re American (or a fan of the US political process), it’s a relevant topic presented well.
Most of us have read this by now. If you haven’t, you need to set that right.
My depth of appreciation for Bayesian analysis is appalling. I love the idea and would like to reach a place where it’s my default method for analysis. I’m not expecting this book to take me all the way there, but so far it’s a splendid, practical introduction.
I haven’t had the opportunity to get into this one in detail. I’m convinced that underneath it all, I’m a Bayesian, so I doubled down on the Bayesian books.
R has dramatically increased what I’m capable of doing with the visual presentation of data. If I ever feel my imagination or enthusiasm for visualization starting to flag, this book picks me up.
I actually hadn’t gotten around to Moneyball until January of last year. Michael Lewis is fairly reliable and this one is a nice, easy read. I saw Billy Beane speak at an event at SAS last summer. They really use this stuff. Beane loves numbers and can talk about them well.