Yesterday, we had a meeting of our EMILE network on statistics for population genetics (in Montpellier) and we were discussing our respective recent advances in ABC model choice. One of our colleagues mentioned the constant request (from referees) to include the post-ABC processing devised by Fagundes et al. in their 2007 ... [Read more...]

“We have seen that a perfect correlation is perfectly linear, so an imperfect correlation will be `imperfectly linear’.” page 128 This book has been written by two linguists, Shravan Vasishth and Michael Broe, in order to teach statistics “in  areas that are traditionally not mathematically demanding” at a deeper level than ... [Read more...]

Here is an interesting question from Tomàs that echoes a lot of related emails: I’m turning to you for advice. I’m facing problem  where parameter space is bounded, e.g. all parameters have to be positive. If in MCMC as proposal distribution I use normal distribution, then ... [Read more...]

Following my earlier posts on the revision of Lack of confidence, here is an interesting outcome from the derivation of the exact marginal likelihood in the Laplace case. Computing the posterior probability of a normal model versus a Laplace model in the normal (gold) and the Laplace (chocolate) settings leads ... [Read more...]

The French agronomy research institute INRA is organising a Fall school in La Rochelle, Nov. 28 – Dec. 02, on Bayesian methods, oriented towards the applications in food sciences, environmental sciences, and biology. The provisional program (in French) is ■ Initiation aux outils informatiques R et WinBUGS (TP et réalisation de projets sur ... [Read more...]

In conjunction with the normal-Laplace comparison mentioned in the most recent post about our lack of confidence in ABC model choice, we have been working on the derivation of the exact Bayes factor and I derived an easy formula for the marginal likelihood in the Laplace case that boils down ... [Read more...]

Mark Girolami sent me this announcement for six PhD studentships in Statistical Methodology and Its Application at University College London (UCL) that are great opportunities for anyone interested in computational statistics! The studentships are attached to the Department of Statistical Science at University College London, and a subset of them ... [Read more...]

The Parco Nazionale Gran Paradiso and the Università di Pavia are organising a summer school on “Advances in species distribution modelling in ecological studies and conservation” in Pavia and Cogne, 12-18 September 2011. This school includes R and Winbugs tutorials, regular classes, plus a field trip to the park, so this ... [Read more...]

I've written before about the up- and downsides of the plyr package -- I love it's simplicity, but it can't be mindlessly applied, no pun intended. This week, I started building a agent-based model for a large population, and I figured I'd use something like a binomial per-timestep birth-death process ... [Read more...]

Yesterday, Andrew Thomas and José Blanchet posted a note on the Bernouilli factory on arXiv. This short paper links with the recent paper of Flegal and Herbei I commented earlier. Considering the special target Thomas and Blanchet develop an elaborate scheme of cascading envelopes that converge to f from above. ... [Read more...]

In search of bin countsI look at histograms and density functions of my data in R on a regular basis. I have some idea of the algorithms behind these, but I've never had any reason to go under the hood until now. Lately, I've been looking using the b... [Read more...]

[Warning: this post about a race I ran yesterday is unlikely to be of interest for most readers!] Following my (un)reasonable time last year, I registered again for the annual “Cross de Bercy”  run by the Sport Club of the Finance Ministry (with whom/which CREST is affiliated). This ... [Read more...]

My neighbour is an half-retired entrepreneur who still runs his electric engine company. A few weekends ago, he came to me with the following physics question related with one of those engines: given a primary disk rotating at the angular speed of ω0 and a secondary disk located on the first ... [Read more...]

Following a link on R-bloggers, I ended up on this page (with a completely useless graph that only contained the pieces of information 5% in 1900 and 55% in 2000). The author (Ralph Keeney) reports on “A remarkable 55 percent of deaths for people age 15 to 64 can be attributed to decisions with readily [...] [Read more...]

Through the webpage of the Advanced Monte Carlo Methods I & II, given a few years ago by Michael Mascagni at ETH Zürich, I found a link to the scanned version of the 1964 book Monte Carlo Methods by Hammersley and Handscomb. This is a short book, with less than 150 pages, ... [Read more...]

The most visited post ever on the ‘Og was In{s}a(ne), my report on Radford Neal’s experiments with speeding up R by using different brackets (the second most populat was Ross Ihaka’s comments, “simply start over and build something better”). I just spotted two new entries ... [Read more...]

“We don’t need likelihood functions; we just need to know how to simulate from [them] (…) We don’t need models with sufficient statistics; we just need summary statistics (…) We don’t need to be Bayesian; we just need to be approximately so. We don’t need theory to tell ... [Read more...]

The 24 questions asked by John Halton in the conclusion of his 1970 survey are Can we obtain a theory of convergence for random variables taking values in Fréchet spaces? Can the study of Monte Carlo estimates in separable Fréchet spaces give a theory of global approximation? When sampling functions, ... [Read more...]

“The only good Monte Carlos are the dead Monte Carlos” (Trotter and Tukey, quoted by Halton) When I presented my [partial] history of MCM methods in Bristol two months ago, at the Julian Besag memorial, Christophe Andrieu mentioned a 1970 SIAM survey by John Halton on A retrospective and prospective survey ... [Read more...]

I received at last my weekend edition of Le Monde and hence the solution proposed by the authors (Cohen and Busser) to the puzzle #14. They obtain a strategy that only requires at most 19 steps. The idea is to start with a first test, which gives a reference score S0, and ... [Read more...]