Congratulations to my friend Ed Paradela for being named as the new Director of Department of Science Region 7. I was assigned to work with Ed during my service in the Peace Corps, and I was also fortunate enough to live with, and learn from, he and his family. Director Burt Llanto, who I also worked with closely set a very high bar, and I know Ed will do a great job moving forward.

*"Imagine something like a game of hopscotch: one must land on each cell in turn, always moving from left to right, finding a path in which the model’s constraints hold. Where there are many cells at the same time, finding a path becomes especially difficult. We must consider every possible permutation of those concurrent cells, which is O(n!). That’s the kind of hopscotch that, even when played by computer, makes one re-evaluate one’s life choices.”*

I just released a new gem, an implementation of the visitor pattern in Ruby named Vampire.

High Scalability has an excellent, accessible article on Little’s Law, and how it fits into the threads vs. event-driven debate.

## Great Stats Reference →

I stumbled across NIST’s Engineering Statistics Handbook, a great applied introduction to exploratory data analysis, and a great alternative to wading through Wikipedia. Need to know what an F-test is and how to interpret it in one page or less? Look no further.

NIST is apparently in the process of updating the examples to use R instead of their somewhat obscure Dataplot software.

## Linear Regression →

Nice discussion of mathematical foundations behind linear regressions.

## Task Lists to Fit a Doctor’s Mind

Rob Lamberts has an interesting article on the Health Care Blog about the steps a physician goes through when deciding the best course of treatment for a patient. He discusses how the simple act of ordering lab results can lead to nine or more repeated steps a doctor has to stay on top of, for each patient they treat.