On 03/02/09 07:11 AM, Johan Jansson wrote: > Bartosz Sawicki wrote: >> Hi all, >> >> Have any of you, some experience with posteriori error estimation in >> Dolfin? I'm playing with adaptive refinement techniques, where proper >> error estimation is essential. >> I try to use form language with discontinous Lagrange elements to >> describe error, but the results are somehow strange. >> I found that in Unicorn there are some error estimation code, but I >> can't find any application of it or any example? Is this code working? >> >> regrd. >> BArtek >> _______________________________________________ >> DOLFIN-dev mailing list >> [email protected] >> http://www.fenics.org/mailman/listinfo/dolfin-dev >> > Hi! > > Yes, the Unicorn error estimation code is working. I assume you're > looking at the Unicorn release, which is almost 1 year old now. There > you can find the dual and residual forms, but the demo problems only > computed residuals, not any error estimates with dual solutions. > > Why not check out the development version of Unicorn from the repository > at: > > hg clone http://www.nada.kth.se/cgi-bin/jjan/hgwebdir.cgi/unicorn-kth > > (look at for example revision 103) > > The ucsolver/icns/cube example has a "run.py" Python script which runs > the full adaptive algorithm with primal and dual solvers and adaptivity > based on the error estimate for drag output. You can use the mesh from > here to test: > > | wget -O mesh.xml http://www.csc.kth.se/~jjan/transfer/cube-mesh11.xml > <http://www.csc.kth.se/%7Ejjan/transfer/cube-mesh11.xml> > > |The residual is computed by multiplication of a DG test functihn as > Anders mentioned. The dual forms are generated by hand, it would be nice > to be able to automate that part as well. >
Thank you for answers. I see that there are many ways to solve the problem, but Dolfin doesn't have any special facilitate. I have written manual integration for error estimation, it works fine. Till Fenics would be able to handle error estimation this using forms in more natural way, I'll keep using my code. cheers, BArtek _______________________________________________ DOLFIN-dev mailing list [email protected] http://www.fenics.org/mailman/listinfo/dolfin-dev
