## Archive for October, 2009

Posted in Kids, Travel with tags , on October 31, 2009 by xi'an

Here is a guest post by Rachel:

# LOOK !

## Bayesian Core: a Madrilean course (2)

Posted in Statistics, Travel, University life with tags , , , on October 31, 2009 by xi'an

The course at UAM was attended by about twenty persons, both students and faculty, which is a fairly good audience in a Master program where most students of UAM opt for an Analysis curriculum, rather than Statistics or Probability. Actually, some students came from La Universidad Complutense de Madrid. (The name always evokes in me some connection with computing, but actually comes from the Latin complutum!) I did not go as far as I wished since I only covered the two first first two chapters of Bayesian Core. (A side consequence of the editing of “Introducing…” is that I now avoid inverting the digit and “first” in a sentence!) But I managed to go through the important points of selling the Bayesian approach to testing as concentrating on model choice and its consequences, and also of presenting the fully automated regression [Bayesian] analysis using n as the g (!) in the g-prior modelling. Those two points will come reinforced in the revision of Bayesian Core, especially since Jim Albert introduced g-priors in his second edition! As an aside, the UAM campus was quite enjoyable due to an Indian summer spell that turned the morning light in the autumnal leaves into a very warm colour.

## Bad graph of the day

Posted in Books, Statistics with tags , , , on October 30, 2009 by xi'an

Another meaningless graph found in the November issue of La Recherche: a histogram of the predictions of the World population by 2005 attached to a brief discussion of the challenges of providing food for this population. No mention is made of the source(s) for this absurd agglomerate of predictions, (could I add mine as well?!) while the discussion picks the median prediction for its reference number: as if Science was run by majority rule… As an unflattering coincidence (for La Recherche!), the other French monthly popular science magazine Pour la Science has simultaneously published a rather well-argumented special issue on randomness (by Jaroslaw Strzalko, Juliusz Grabski and Tomasz Kapitaniak who are Polish physicists), refering to one recent paper by Persi Diaconis on the randomness of coin tosses. Being associated with Scientific American certainly helps in producing quality papers! (There is also a paper by Ivar Ekeland in the same issue, as well as the paper by Andrew Gelman already signaled.)

## Before they are hanged

Posted in Books with tags , , , on October 29, 2009 by xi'an

The second tome of Joe Abercrombie’s The First Law trilogy, Before they are hanged, does not leave live up to its expectations. I really enjoyed the first volume, The Blade Itself, overreacting a little bit at the novelty in tone and style by suggesting a ranking in the New Yorker fantasy list, but this second volume definitely falls very short from my expectations! I mentioned the potential connections with David Eddings in my earlier post, and they are much more apparent in the second volume: the characters are lacking depth but not stereotypes, the way the central society behaves is caricatural, the enemies have endless facilities and power… The (obligatory) quest has started but the crossing of the continent to the other sea is very boring and does not bring further perspectives on the six characters, some of which like Qai and Longfoot (!) do not even seem to have a proper role in the trip. They were interesting anti-heros in the first volume, they are no longer that interesting nor that anti-heros. The other main character, Glokta, has also lost some of his cruel appeal by playing the rescuer to all distressed maidens crossing his path. Overall, the witty exchanges found in the first volume have mostly disappeared from the second (or I do not find them so funny!) So I fear this is another example of a good start turned into a tepid follow-up (please let not this happen to The Name of Wind!). Given the mixed reviews on the third volume, I am not sure I will give it a try!

## Creationism in UK schools?!

Posted in Kids, Travel with tags , , , on October 28, 2009 by xi'an

“It was found that Britons were almost three times more likely than Egyptians to want creationism and intelligent design to be included in the teaching of evolution.” The Guardian, October 25, 2009

I was reading the Guardian on the flight to Madrid and there was this terrible statistics that 54% of the UK public wanted creationism to be “taught” in public schools. This is worse than in the US… Quite an appaling statistic for the Darwin year and Darwin’s country! Of course, this kind of result should be taken with a pinch of salt since the same Guardian reports that four Britons out of five repudiate creationism, along with a paraodical region-by-region map where each region has more than 25% of the people in favour of creationism or “intelligent” design! (The quote above is rather dumb as well! Why should Egyptians be more in favour of creationism?!) Here is a blog reporting more clearly on a similar study, conducted by Science, with France coming almost on top of Darwinian countries! Except that Britain is fairly close to the top as well. So this may end up being a catchy title making too much of a limited or poorly conducted survey. To conclude that “More than half of British adults think that intelligent design and creationism should be taught alongside evolution in schoolscience lessons” based on 973 Britons is extrapolation to the third power…

## Verbatim code

Posted in Books, Linux with tags , , , , , on October 27, 2009 by xi'an

When rendering computer code and output in LateX, I usually resort to the basic verbatim environment, as in

\begin{verbatim} > matrix(1:4,ncol=3) [,1] [,2] [,3] [1,]    1    3    1 [2,]    2    4    2 Warning message: data length [4] is not a submultiple or multiple of the number of columns [3] in matrix in: matrix(1:4, ncol = 3) \end{verbatim}

However, I do not like the way tildes are represented in this environment, being considered as accents rather than standard characters, as in the R command

lm(y~x)

Asking Springer-Verlag TeXperts for help during my last round of corrections of Introducing Monte Carlo Methods with R before it left for the printer, I received a solution which is to use the alltt environment instead, which allows for LaTeX commands within the verbatim rendering, using $$\sim$$ instead of ~. (I have been looking for this solution for ages and of course, provided with the solution of using alltt instead, I immediately found a lot of related posts!) There is a drag, though, which is that curly brackets {} vanish from the R code, unless replaced with the LaTeX versions $$\{$$ and $$\}$$.

When ~is used in the in-line rendering

\verb+lm(y~x)+

alltt does not work but a simple

\verb+lm(y+$\sim$\verb+x)+

is sufficient.

## Bayesian Core: a Madrilean poster

Posted in Statistics, Travel, University life with tags , , , , , , , , , on October 26, 2009 by xi'an

The Departamento de Matemáticas de la UAM has had a really nice poster made for my course! Further details are

Días: 28 y 30 de octubre de 2009
Horario: de 10 a 13 h.
Lugar: Facultad de Ciencias, módulo 16, aula 420