Introducing StataStan

stan logo

Thanks to Robert Grant, we now have a Stata interface! For more details, see:

Jonah and Ben have already kicked the tires, and it works. We’ll be working on it more as time goes on as part of our Institute of Education Sciences grant (turns out education researchers use a lot of Stata).

We welcome feedback, either on the Stan users list or on Robert’s blog post. Please don’t leave comments about StataStan here — I don’t want to either close comments for this post or hijack Robert’s traffic.

Thanks, Robert!

P.S. Yes, we know that Stata released its own Bayesian analysis package, which even provides a way to program your own Bayesian models. Their language doesn’t look very flexible, and the MCMC sampler is based on Metropolis and Gibbs, so we’re not too worried about the competition on hard problems.

2 thoughts on “Introducing StataStan

  1. You’re welcome. I’d like to do some proper comparison of Stata 14 with John Thompson’s Stata-to-BUGS commands and StataStan. It’s early days for Stata’s Bayesian department but it’s nice anyway to see mainstream software taking it on.

  2. I’m late to this, but I just looked at Stata’s own Bayesian inference package. As far as I can tell, it only ever runs a single chain!!! All the convergence diagnostics shown in the 267-page manual are things like traceplots…on only one chain. There’s one note on pg 15 of the manual suggesting that some people (Gelman and Rubin 1992) suggest running multiple chains to check convergence, but it seems as if the Stata implementation can’t actually do this. I know there are some folks (Geyer??) who argue that you don’t need multiple chains to assess convergence, but I don’t think that’s a common view anymore. It seems almost irresponsible to release software in this state. In general I really respect Stata, because they usually seem to take the time to get things right, but this Bayesian package feels like a misstep or, at best, an “alpha” release.

Leave a Reply

Your email address will not be published. Required fields are marked *