Travail de recherche/Working paper
Résumé : This paper presents a Markov-Chain-Monte-Carlo (MCMC) procedure to sample uniformly from the collection of datasets that satisfy some revealed preference test. The MCMC combines a Gibbs-sampler with a simple hit and run step. It is shown that the MCMC has the uniform distribution as its unique invariant distribution and that it converges to this distribution at an exponential rate.