Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • StataStan command to fit Bayesian models with Stan (Hamiltonian Monte Carlo etc)

    The alpha-testing version of StataStan is now online at https://github.com/stan-dev/statastan/tree/alpha-test with blog post here: https://robertgrantstats.wordpress.c...ing-statastan/

    Stan (mc-stan.org) gives you access to Hamiltonian Monte Carlo, a big advance in stability and speed over the Metropolis-Hasting and Gibbs methods available in Stata 14, BUGS and JAGS. We would particularly like to know that it runs on various Stata versions, flavors and operating systems, so please try it out and give feedback either on GitHub, or here, or just by emailing me at [email protected]

    It has mostly been worked up in Debian and then tested in Windows V*sta with versions 11 and 13, IC and SE.

    Robert on behalf of the Stan developers

    PS It's just possible I might be talking about this at the UK SUG in September.

  • #2
    Robert, good day to you. Please where and how can I download stan in stata.

    Comment


    • #3
      Hi Olayiwola
      You can do this:
      Install CmdStan (see mc-stan.org)
      Then, inside Stata: -ssc install stan-
      if you are using Windows, -ssc install windowsmonitor-
      But if you are using Stata 16, you might prefer to go through the Python interface. I will write some more on this when time allows, at the StataStan page and here.

      Robert

      Comment

      Working...
      X