nerd sniped: books about distributions

It was decided today that I need to learn more about the Poisson distribution, and preferably not just from Wikipedia.  Thus, I decided a two-pronged approach to build up my intuition: playing with it in R and reading up on it in Johnson et al.'s Univariate Discrete Distributions.

What should have been a quick trip to the library ended up involving me sitting on the floor (bottom shelves always make me stay longer for this very reason) and browsing the books for at least ten, maybe fifteen minutes.  That's not terribly long, you might say, but I'm an incredibly decisive person* and this is a topic that most normal people would spend fifteen minutes avoiding.

{* Aside: As long as I'm the only one making the decision, I'm very decisive.  As soon as other people's opinions come into play, I'm wishy-washy like whoa.  Anecdote: my wedding invitations were selected down to font and ink color probably within 7 minutes of opening the two huge binders full of options.  But, ask me for restaurant preferences for a group dinner and you'll get an annoyingly placatory response. }

So I end up leaving with not just the book that I had intended to check out, but also Severini's Elements of Distribution Theory and Consul's Generalized Poisson Distributions.  Great reading for a Friday night.

While I was there, I also stumbled upon a book called A Folio of Distributions, which, as far as I could tell, consisted entirely of plots for all of its 500-some pages.  It was originally published in 1987, long before the age when a student of statistics could simply fire up R in order to see how a distribution behaves in different contexts.  I almost checked it out, but given the facts that A) I was on foot and B) the computer I was already carrying could produce the same results, I let it be.

The moral of the story: I love university libraries, especially those belonging to absurdly privileged institutions.  If these books don't grow my brain, they'll certainly grow my muscles.

No comments: