It seems as good a time as any to link to this chestnut from McSweeney’s. Warning : full of highly profane language.

Did you know Krish Eswaran has a blog? Neither did I, until today. He appears to not be updating it, however. C’mon Krish, more posts!

Andrew Gelman pins down one of the things that annoys me about arguments based on personal finance — it’s not true that we do things for money or for fun, unless “fun” is really broadly construed. Plus there’s this zinger at the end: “as a statistician, I’m not impressed with an argument when it doesn’t work on the example it’s been applied to.”

A pretty cool video on hand-pulled noodles.

More chaconne than you can handle.

Via CT, an amazing cartoon in which Donald Duck meets Glenn Beck.

How to write about Pakistan, inspired by the classic How to write about Africa.


Definitely not the desired interface

I am writing this from my new iphone, for which there is a wordpress app that will let me write posts from my phone. I have to say that although the concept is cool, having to write anything complicated on this touch screen is a little maddening. That being said, I hope to blog soon about some recent reads, canonical angles between subspaces, and a generalization of Fano’s inequality that I learned about in a paper by Bin Yu.

LaTeX instructions

To use \LaTeX on this blog, you first use a dollar sign and “latex” like this:

$latex x = \frac{ - b \pm \sqrt{b^2 - 4 a c} }{2 a}

then you can close it with another dollar sign:


Putting it all together, the renderer will give you:

x = \frac{ - b \pm \sqrt{b^2 - 4 a c} }{2 a}

Basically, it’s like writing latex, except that every time you have to go into math mode, you have to type “latex” right after the first dollar sign.

pardon my dust

This blog is going down for a day or two (or perhaps the weekend) in order to move to a new home hosted by They support LaTeX and I’ve decided to move the thing over there. It’ll take a few days for the domain redirection to catch up, so be forewarned.

The blog has now moved, and here is a test of the new embedded \LaTeX features:

I( X \wedge Y ) = \sum_{x,y} P(x,y) \log \frac{P(x,y)}{P(x) P(y)}


I(P,V) = \sum_{x,y} P(x) V(y | x) \log \frac{ V(y | x) }{P(y)}