jackman.stanford.edu/blog
• 114th U.S. Senate
• ideal point estimates svg pdf csv 10/1/15
• scatterplot against 2012 Obama vote share svg pdf
• roll call object: RData
• 114th U.S. House
• ideal point estimates svg pdf csv 10/1/15
• scatterplot vs Obama vote share svg pdf
• roll call object: RData
• Bayesian Analysis for the Social Sciences Wiley; Amazon; errata as of 3/6/15

## Saturday October 30, 2010

Filed under: Australian Politics — jackman @ 5:32 am

610 picograms of dioxin per gram of sediment in Homebush Bay; 350 picograms/g in the Parramatta River and east (towards the bridge/ocean) at Rozelle. In a “relatively pristine estuarine environment” you’d get 2.3 pg/g.

I have no real idea how to interpret these numbers. A picogram is one trillion-th or 10^-(12) of a gram. The US EPA (via Wiki) said in 2007 that the typical American is injesting 119pg of dioxins per day, with beef and dairy products the largest sources. Some work examining the link between exposure levels and cancer prevalence appears here.

According to the SMH:

large tracts of sediments in the harbour are more contaminated with dioxins than Tokyo Bay or New York Harbour, and are among the most contaminated in the world.

Further:

The contamination has a ”signature” that matches the cocktail of dioxins produced by Union Carbide during manufacture of the now-banned pesticide 2,4,5-T and the defoliant, Agent Orange, used during the Vietnam War.

I wonder about actually dumping sand or dirt something inert to help smother this stuff on the bed of the Harbour? Or are we talking about just too large an area for that to be practical? I’m thinking of something that might do a reasonable job of acting like “naturally occurring” sediments, burying the bad stuff, but just arriving more quickly than it would otherwise.

Comments Off on thanks Union Carbide

## Thursday October 28, 2010

Filed under: flight nerdery — jackman @ 3:30 pm

Earlier today I flew SFO to Chicago. I got a window seat on the left side of the plane, which gave me a great view of the field as we were on taxi to 01R for departure. I was also listening in on Channel 9 (thanks United) and listened carefully after I saw a Continental execute a go-around after visibly bouncing back into the air (at least once) while attempting to land on the 28s. What I heard went something like this:

Continental: Continental executing go-around.
TWR: Roger Continental, fly heading two eight zero, three thousand.
Continental: Roger, two eight zero, three thousand. Continental.
TWR: Say reason for the go-around.
Continental: Bounced landing.
TWR: Say again Continental?
Continental: Bounced… Landing.
TWR: Roger Continental.
Unidentified source: “Bounce…!”

Cold…! Reeks of military style rookie hazing. Nice.

Comments Off on “Bounce…”

## Saturday October 23, 2010

Filed under: computing,R,statistics — jackman @ 10:35 pm

I gave an informal talk on my Mac based “workflow” at Stanford on Friday. Â I talked a lot about Aquamacs as a tool for editing LaTeX (I use MacTeX) and for working with R (thanks auctex and ess, respectively). Â Skim also got a mention; I emphasized TeX-PDF synchronization.

Some of the students were asking for my Aquamacs customizations. Â Here is the current version of

~/Library/Preferences/Aquamacs Emacs/customizations.el

from my laptop machine. Â There are some definitions that are specific to my machine, but otherwise it is reasonably straightforward.

I’d appreciate any discussion on this â€”things I’m missing, things I’m doing in a very old fashioned way etc.

## Monday October 18, 2010

Filed under: flight nerdery — jackman @ 2:51 pm

About time! Apparently Air New Zealand is considering this for long-hauls like LAX-AKL etc.

Beginning next year, money-conscious long-haul travelers can take “cuddle class” and book a row of seats that convert into a couch or bed.

Comments Off on cuddle class

## Wednesday October 13, 2010

Filed under: R — jackman @ 3:29 pm

R/SQL/scripting, oodles of data, a willing outlet for write-ups. Any takers?

Do some good. Intern at Barron’s, the New York financial publication with a decades-long tradition of investigative journalism and a more recent commitment to data analytic exposure of fraud in finance, business and healthcare. Bring us your zeal and your data munging skills and we’ll teach you how to be a writer, stock analyst and gumshoe investigator. We assemble unique data sets to scrutinize activities like Jim Cramer’s CNBC stock picking, for-profit colleges, Medicare rip-offs and the hyping of US-listed China stocks. Our stories achieve instant reform of bad situations. Please help us make Wall Street safe for the little guy.

We will hire one programmer-journalist intern to work from June through August of 2011. Desirable skills would include programming fluency in R (for data analysis and visualization), a scripting language (like Perl or Python) and SQL. If you can write, you will also get stories published under your byline.

As part of the Dow Jones News Fund, the program provides free pre-internship training seminars, 10 weeks’ salary, $1,000 scholarships for interns returning to school, and a travel allowance. The salary is not princely, so it helps if you have a place to stay in the New York area. APPLICATION DEADLINE: November 1, 2010 Apply online at https://www.newsfund.org…Click Programs…then Internship Applications…make sure to complete the business reporting form. Also, send a resume, writing and code samples to me — Bill Alpert — at william.alpert@barrons.com and call me with any questions at 212.416.2742. ### Commbank would only lose 0.2% in a mortgage meltdown Filed under: general — jackman @ 4:16 am Strains credulity. From the SMH: Commonwealth Bank, which did its own stress test, said last month that under a â€œhigh-stress scenario,â€ with interest rates at 14 per cent, unemployment at 10 per cent and property values down 30 per cent, it would see losses of$740 million, or 0.2 per cent of its total book, not including additional insured losses.

One can’t but wonder about the profits being made when interest rates go to 14%, new lending dries up (so fewer capital raising costs).

The more interesting scenario is the one like in the US: the cash rate close to zero, unemployment at 10% for a prolonged period, negative equity situations all over the place.

## Thursday October 7, 2010

Filed under: type — jackman @ 11:54 pm

The new Gap logo. Freaking ridiculous. I am sure they tested the shit out of it, but bold Helvetica…? In 2010?

Join Apple Ipad/iPhone (which has actually switched to Helvetica Neue), American Airlines, American Apparel, IRS forms, Target, Crate and Barrel, YouGov (which I think is actually Arial) etc etc, making a bland brand all that more so…

## Wednesday October 6, 2010

Filed under: general — jackman @ 4:30 pm

Give me an incentive to go paperless.

Right now I can’t help but look at your “green initiative” and see — first and foremost — more profit for you. Will you charge me fewer account fees if I go paperless? You won’t be printing and mailing me statements every month etc.

Sure, we’re both better off if the planet is greener, but those benefits are really quite diffuse. Meanwhile, you get a direct benefit from me (us?) doing something for the planet, but I don’t. You will no doubt trumpet to your investors what good you are doing for the planet while extracting more profit from me.

Your reminders that “there is no cost to going paperless” are really quite lovely. But I’d like to share the savings with lower account fees, thanks.

$3 for a 3 working day transfer to a brokerage account.$12 for processing a foreign currency transfer. And you offer me warm inner glow for going paperless.

Try harder.

### How fast is JAGS?

Filed under: computing,R,statistics — jackman @ 1:56 am

From Martyn Plummer, on the JAGS news blog. Key graph below, showing a few outlying cases in which JAGS is substantially slower than OpenBUGS, but generally, JAGS performs quite favorably.

Key point from Martyn:

Incidentally, these figures are for JAGS with the glm module loaded. The glm module is not loaded by default. If you are not using it for generalized linear mixed models then you might be missing out.

Comments Off on How fast is JAGS?

## Saturday October 2, 2010

Filed under: Australian Politics,statistics — jackman @ 11:19 am

Centrebet is offering the following prices on the date of the next Australian Federal election:

The fact that the 2011 election is the market favorite is being a little over-interpreted in some quarters. The probability of a 2011 election implied by these price is .35. Quite simply, a 2011 election is not an “even-money” proposition (or better).

Interestingly, Centrebet’s overround with these prices is 1.34, whereas in the two-horse election markets the overround was usually way lower, around 1.06. “Come in spinner…” and all that.

The best (non-technical!) definition of the over-round I’ve found is at bettingmarket.com:

“The over-round (Vigorish, or simply “vig”, or “juice”) is what gives the bookmaker his profit, his sun tan and his cigars.”

Next Page »