[RMCProfile-users] What is a reasonable acceptance rate?

[Tucker, Matthew G.]

Dear Karl,

When the fit has converged and equilibrium reach you would ideally like ~50% of moves to be accepted. 

There are various caveats of course, if you are swapping atoms with translations and you are using closes approach constraints with different values for the various atomic sites, then eventually the number of swaps being accepted may fall close to zero, also if any of the data sets are weighted heavily then this will lower the number of accepted moves at equilibrium. 

So in general if you have 25 to 50% of accepted moves at equilibrium, it is okay and you can collected different configuration after running for a reasonable number of moves if you want to improve statistic on a particular distribution. 

Hope this helps, if not let me know and I will try again.

Best wishes,
Matt

On 6/8/16, 1:07 PM, "rmcprofile-users on behalf of Karl Pazdernik" <rmcprofile-users-bounces at rmcprofile.org on behalf of ktpazder at ncsu.edu> wrote:

>Hello experts,
>
>When running the RMC algorithm, I assume initial moves are almost always
>accepted.  Once the [image: \chi^2] value begins to converge, though, is
>there a reasonable acceptance rate?  That is to say, if you were to tally
>the number of random moves accepted and divide by the iteration number,
>should this proportion converge to a particular number (range of numbers)?
>
>
>-- 
>Karl Pazdernik, Ph.D.
>Department of Statistics
>North Carolina State University
>-------------- next part --------------
>An HTML attachment was scrubbed...
>URL: <http://lists.rmcprofile.org/pipermail/rmcprofile-users/attachments/20160608/b6708c2d/attachment.html>
>_______________________________________________
>rmcprofile-users mailing list
>rmcprofile-users at rmcprofile.org
>http://lists.rmcprofile.org/mailman/listinfo/rmcprofile-users
>