rencontres R à Lyon, 27-28 juin 2013

Simon Barthelmé just sent me the announcement that the next French R conference (Rencontres R) will be in Lyon, on June 27-28, 2013, and that the call for propositions is now open. (No possible confusion with MCMSki IV: the later is not in Lyon but Chamonix, not in June but January, and not only about R—even though there will be a session!—but about all computational aspects of statistics, well mostly MCMC!)

Trip to Lyon

This was my first trip to Lyon in about… 35 years, I think, but I did not have much time to tour the city! My original plan was to go climbing with Ivan near La Meije right after the talk, but our respective knees were hurting for the past week at least (since Utah in my case…), so we decided to postpone. I think the talk itself went on rather ok (?!), with good questions for possible extensions at the end. And Martyn Plummer (the creator and maintainer of JAGS) came with a most relevant question about the difficult comparison with completely independent MCMC chains, a point Colin also raised a while ago.

Séminaire à Lyon

This week, I am going to Lyon to give a seminar in Université Lyon I. I will talk about the recent Rao-Blackwellisation papers we wrote, hopefully managing to make sense with both statisticians and probabilists, unlike in recent talks of mine in probability seminars… The slides are close to those I presented in Wharton last month, except for recent additions on parallelising random walk Metropolis-Hastings algorithms as well.