Thursday, October 31, 2013

Climate Science Resource

I found this resource page on Skeptical Science through the NY Skeptics podcast.

The nice part about the way the page is organized is that it tries to debunk "climate change denial" claims at two levels: a basic level that you could use to engage middle school kids (simplified explanations), and an intermediate-level that is like a scientific paper, and backed by more data, references, and nuance.

You should check it out.

Monday, October 21, 2013

Review: Land of the Seven Rivers

I read Sanjeev Sanyal's "Land of the Seven Rivers: A Brief History of India's Geography" with much interest after I heard a podcast of his talk at LSE.

It presents Indian history (and prehistory) through the lens of geography and climate. The new perspective allows for some fascinating reinterpretations.

For example, Sanyal claims that the Ramayana was a oriented along a North-South axis (Ayodhya to Sri Lanka), and that the Mahabharata, was more aligned along an approximately East-West axis. Remnants of these paths still exist in the form of segments of national highways.

Similarly, he tells the fascinating story of why Gautam Buddha went to Sarnath to propagate his new philosophy (it sits at the intersection of two important commercial routes, which allows for rapid dissemination of goods and ideas). He dwells on the role of "pillars", and how some of them became a canvas on which successive waves of emperors would carve out their names for posterity.

One surprising aspect (which perhaps should not have been that surprsing) is how much the climate has changed even in the last few thousand years, and how a response to such massive climate change has shaped the course of history (example, the end of the Indus-Saraswati valley civilization). As we deal with a climate crisis of our own, it is useful to look at the massive disruption such upheavals cause.

If another reason to study history is to learn how we got where we are, then I think this book shows us how big of a role Nature and fortune (as opposed to politics and monarchs, which are the usual lens through which history is told) played in shaping the course of events.

The book is full of factoids and clever observation of patterns. Throughout the book there is an underlying concept of "continuity" and the idea of "nationhood" extending from 90 million years ago, through the merger of the Indian subcontinent with Asia, the early river valley civilizations, through the last couple of millennium, up to the present day with a discussion of the Indian diaspora.

It is truly a fascinating read, and very highly recommended. 

Saturday, October 12, 2013

Nicer Octave Plots for Presentations/Documents

GNU Octave has a very capable plotting system (which is transitioning away from gnuplot as its backend, even as we "speak"). Quite often, you generate a figure and make it look like you want:

You then want to incorporate it into some other document - say a presentation or a document.

So you say:

print -dpng test0.png

Note that you could also use -depsc2 or -dpdf to generate EPS or PDF formats among a whole host of other possibilities.

You get a figure that looks like:

Clearly imperfect.

There are a number of ways to improve the image.

These include writing to tikz, TeX, which gives you exquisite control over all features, but this can take too much effort.

Or you could spend time fiddling with adjustable parameters to make the plot look just right. If you are trying to work on "the" central plot of a paper, or a PhD thesis, then such care is certainly warranted.

But if you are merely dissatisfied (as I am) with the size and font of the text, there is an easy fix.

print -dpng -FHelvetica:18 test1.png

Sunday, October 6, 2013


As I mentioned previously, one of my colleagues is running an MCMC seminar this semester, which is essentially a journal club that is trying to read and discuss some of the more seminal papers in the field.

The second paper we discussed was Hastings famous 1970 paper, which generalized the "Metropolis" algorithm and couched it in a more general form (the Metropolis-Hastings algorithm).

The story behind this man and his famous paper turns out to be every bit as fascinating as that behind the original paper.

From 1966 to 1971, Hastings was an Associate Professor in the Department of Mathematics at the University of Toronto. During this period, he wrote the famous paper listed above (which generalised the work of N. Metropolis, A. Rosenbluth, M. Rosenbluth, A. Teller, and E. Teller (1953), "Equations of state calculations by fast computing machines", J. Chem. Phys. 21, 1087-1091). Hastings explains: 
When I returned to the University of Toronto, after my time at Bell Labs, I focused on Monte Carlo methods and at first on methods of sampling from probability distributions with no particular area of application in mind. [University of Toronto Chemistry professor] John Valleau and his associates consulted me concerning their work. They were using Metropolis's method to estimate the mean energy of a system of particles in a defined potential field. With 6 coordinates per particle, a system of just 100 particles involved a dimension of 600. When I learned how easy it was to generate samples from high dimensional distributions using Markov chains, I realised how important this was for Statistics, and I devoted all my time to this method and its variants which resulted in the 1970 paper.
He seems to have been a very fascinating character.

He wrote only three peer-reviewed papers all this life, and supervised only a single PhD student. 

Wednesday, October 2, 2013


1. Matt Walsh writes a funny and compelling post on how not to be an ass when other people's kids get cranky in public (h/t Julia).
I had an older guy complain to me recently about babies that cry during church. He said: “Back when our children were babies, you didn’t have this problem.” Interesting. Apparently babies didn’t cry in the 50′s. The whole “crying baby” thing is a new fad, it would seem. These folks who had kids a long time ago seem to have a rather selective memory when it comes to their own days of parenting young kids. They also tend to dismiss the fact that modern parenting presents unique challenges, some of which didn’t apply several decades ago. I always love the older folks who lecture about how THEIR kids weren’t as “attached to electronics” as kids are nowadays. That’s probably true, but mainly because, well, YOU DIDN’T HAVE ELECTRONICS. You had a toaster and a black and white TV with 2 channels, both of which were pretty easy to regulate. But, sure, congratulations for not letting your kids use things that didn’t exist. On that note, I have a strict “no time machines or hover-boards” policy in my home. It is stringently enforced. I’m thinking of writing a parenting book: “How to Stop Your Child From Becoming Dependent Upon Technology That Isn’t Invented Yet”
2. Putting time into perspective: Like the "pale blue dot", there is something liberating about looking at ourselves from afar. It brings the relative insignificance of our personal (or planetary) struggles into sharp relief.