Sascha Yes, a similar method (interpolation instead of evaluation, possibly after adding more spectral coefficients) was implemented by Marcus Ansorg in the TwoPunctures thorn as well.
-erik On Wed, Mar 2, 2016 at 10:42 AM, Sascha Husa <[email protected]> wrote: > On Wed, Mar 2, 2016 at 4:23 PM, Erik Schnetter <[email protected]> wrote: >> Alfred >> >> Solving the equations is fast, the slow part is evaluating the >> solution on the finite differencing grid. If you want to use the same >> grid multiple times you can checkpoint the initial conditions. If you >> want to use different grids, then you'll have to repeat the slow part >> anyway. > > Hi Erik and all, > > some of the time we will just want to run some tests without great > needs in accuracy. > Also, for some extreme initial data the spectral solve may already > take quite long. > > We may look into how the spectral solution is evaluated afterwards on > the finite difference grid, and compare with BAM. > What we routinely use in BAM is some high order interpolation, which > we found to be reasonably fast and sufficiently accurate. > > cheers, > -sascha > > >> There are ways to speed this up by sacrificing accuracy. I don't think >> this makes sense if you want to obtain accurate waveforms, though. >> >> -erik >> >> On Wed, Mar 2, 2016 at 9:32 AM, Alfred Castro <[email protected]> >> wrote: >>> Hi, >>> >>> I am a master student at UIB, and starting to use the Einstein Toolkit >>> for binary black hole evolutions. >>> >>> I am using the TwoPunctures thorn for solving the constraints. In >>> order to speed up tests of grid structures, I would like to just solve >>> on the spectral grid once, store the data in a file, and reread them >>> for different evolutions (e.g. with different finite difference >>> grids). It looks like this is not supported right now. Has anybody >>> done this already? >>> >>> Thanks, >>> >>> -- >>> Alfred >>> >>> _______________________________________________ >>> Users mailing list >>> [email protected] >>> http://lists.einsteintoolkit.org/mailman/listinfo/users >>> >> >> >> >> -- >> Erik Schnetter <[email protected]> >> http://www.perimeterinstitute.ca/personal/eschnetter/ >> _______________________________________________ >> Users mailing list >> [email protected] >> http://lists.einsteintoolkit.org/mailman/listinfo/users -- Erik Schnetter <[email protected]> http://www.perimeterinstitute.ca/personal/eschnetter/ _______________________________________________ Users mailing list [email protected] http://lists.einsteintoolkit.org/mailman/listinfo/users
