R Tutorial: A Bayesian Estimate of Proportion

This is an old chestnut in Bayesian statistics, using the conjugate beta prior to find a beta posterior distribution for a proportion.  If you’re unfamiliar with the calculation of the posterior distribution, there’s a link in the tutorial.



R Tutorial: Teasing Out a Markov Chain

Azzalini and Bowman’s Old Faithful geyser data provides fodder for a lot of data exploration in R (scatterplots, ggplot2, simple regression, kmeans clustering, and Markov chain estimation).  All the really interesting stuff in the tutorial  happens if you click through to Analysis > Models > Standardized Cluster Model.  (The standardized clustering approach is not given in the original paper.)


Beginning: R Tutorials

After a long, slow start, R is catching on with statisticians and (some) scientists at UTSA.  The Biology Department has asked that I use R in teaching biostatistics, and many of the courses for statistics majors are using R rather than SAS (a UTSA tradition).  Students have not been idle; the statistics club has asked me to present an occasional series of R tutorials to get their members up to speed.  Here are the first two tutorials:

These tutorials are all HTML files, generated with RMarkdown.  Students who attend the presentations are also provided with the markdown source files, so they can tweak the code during the presentation.


New Tricks for this Old Dog

Udacity is offering an introductory statistics course this summer, beginning June 25th.  I’ve enrolled, to see how the Big Boys do it.  This is going to put a lot of pressure on traditional universities–especially here in Texas, where we’re busily hammering out the $10,000 Bachelor’s degree.  I figure if I don’t get up with the leaders of the buffalo herd, I’m gonna get trampled or left behind.

Tip from Meep at the Conservative Commune.


A fair 3-way choice using coin tosses

I’d like to make a fair and random choice among 3 alternatives, but the only randomizing device I have available is a coin to toss.  Worse yet, I suspect the coin may be biased.  What to do?


Heretics! Burn them!

Just as UTSA begins to ramp up its Quantitative Scholarship program to inject mathematical reasoning into every crevice of our curriculum, heretics are beginning to doubt the whole enterprise.  Part of the problem is that most remedial and math literacy programs (and textbooks) are filled with bullshit applications and examples (“the rate at which the fluid level in a martini glass will go down”, etc.) that suggest the authors never worked an honest lick in their lives.

Very few people need calculus, but darn near everyone could benefit from knowing the basics of things like the critical path method — planning a move, a wedding, or a family reunion is a major scheduling problem!  But we don’t let the quants who know how to make money teach the Great Mathematically Unwashed.   Instead we stick ’em in Developmental Math 001 and let teaching assistants torture ’em with (pre)calculus.  Then we end up with citizens who get gypped on loans and mortgages, believe in magical thinking about economics, and plan for retirement with lotto tickets.  Nice contribution, State U.

I’m going to be really busy the next few years dishing up useful math and statistics that my students will remember long after their final exams.  And I expect the Math Mafia to give me no small ration of sh*t for doing so.

Update (5 November)Another sermon to the choir from the Math Mafia.


A short tale about the long tail

Lingustics Log has a nice post about early papers on long tail distributions. Good dissertation material, thanks guys!