John Myles White

Recently a friend asked me to help her learn enough math to take the GRE’s.

Now that I do a great deal of my programming in Ruby, there are two tools I constantly miss from my Perl days: Perl::Tidy and Perl::Critic.

If only everything in the world looked as good in photographs as the flowers in the United States Botantical Gardens:

Every program designed to organize files – iTunes, Papers, etc. – would benefit from making the MD5 hashes of every file available to advanced users as another column.

Because of Andrew Gelman’s strong, repeated recommendations, I’ve been reading “Forecasting Presidential Elections” by Steven J.

My advice to teachers of mathematics: explain the intuition behind every theorem.

I always find January a strange month, because the weather tends to get colder over the course of the month, even though the days get progressively longer.

While thinking about linear regression today, I believe that I’ve realized why using clustered sampling and two point average slope calculation works better than least squares regression with scattered sampling.

To follow up on my post earlier today on two approaches to linear model fitting, I decided to do some Monte Carlo simulations to test the relative strength of my two proposals for data collection strategies.

Lately I’ve been thinking about the optimal strategy for data collection when you plan to run a linear regression.

For literally years I’ve been meaning to write a post about my favorite programs and utilities for the Mac, but I’ve always managed to put it off.

As I was finishing reading “The Cult of Statistical Significance” yesterday, the following passage struck me as particularly important: Rothman computed a p-value function – a continuous function of p-values mapped against a range of effect sizes.

Continuing on with my exploration of the National Survey of Drug Use and Health, I thought that I should calculate some simple conditional frequency statistics.

Lately, I’ve been exploring the data set that was recently released by the National Survey of Drug Use and Health.

In our days, serious arguments have been made from data.

To say it once and for all: using " as the escape character for ’ within single-quoted strings is a poor decision in language design.

Equations force a model to be precise, complete, and self-consistent, and they allow its full implications to be worked out.

Today it snowed in Princeton.

Until today I’ve felt that, at this point, Leopard has been patched enough to be relatively bug free.

I don’t care how many movies are available to me.

Today I started reading the Ruby Snips website, which has a pretty good sample of interesting snippets of Ruby code on it.

Here are some more reviews of iPhone apps I’ve had a chance to use.

Here are some very brief reviews of the iPhone 2.0 apps I’ve downloaded and used since getting the leaked firmware last night: Advent: As terrible as Adventure ever has been.

Experiment with dope (as in dopamine).

The bug I complained about on January 12, 2008 is still present in the recently updated iPhone firmware, version 1.1.3.

One of my main frustrations with the 1.1.2 firmware on the iPhone is that the ability to select from multiple e-mail addresses for a single contact while composing a message is an illusion: the e-mail is always sent to the e-mail currently highlighted when