This is the mail archive of the
gsl-discuss@sourceware.cygnus.com
mailing list for the GSL project.
Re: gsl_siman_solve
- To: Fabian Bastin <fbas at math dot fundp dot ac dot be>
- Subject: Re: gsl_siman_solve
- From: Brian Gough <bjg at network-theory dot co dot uk>
- Date: Tue, 18 Apr 2000 12:19:49 +0100 (BST)
- Cc: gsl-discuss at sourceware dot cygnus dot com
- References: <38F5C72C.1308095D@math.fundp.ac.be>
Fabian Bastin writes:
> I have a little question concerning the implementation of
> gsl_siman_solve.
>
> The variable n_evals is used to count the number of evaluations of Ef,
> but n_evals is incremented only after the computation of new_E, not
> after after the computation of E in the first line of the while loop.
> Can you explain why you don't take it into account?
I don't know the reasoning there. Rewriting the simulated annealing
code to use an initialize-update-test interface is on the TODO list
and this will allow the user to keep track of the full
evaluation/accept/reject history.