> On Jul 24, 2017, at 3:26 AM, Lisandro Dalcin <[email protected]> wrote: > > On 23 July 2017 at 21:42, Barry Smith <[email protected]> wrote: >> >> Then explicit methods could automatically use a solve to apply the mass >> matrix inverse for each "right hand side" application and implicit methods >> could automatically "drop in" the mass matrix with the shift when a right >> hand side Jacobian is given also. Instead of requiring the user to handle >> this themselves. Are there any gotcha's that I am missing or is this >> relatively straightforward? >> > > Are the user expected to provide a Mat object, or a callback routine > that assembles a Mat object?
I guess we can support the general but if the user provided a Mat an NO callback I would have the code treat the mass matrix as constant for all time which is a fairly common case. > For generality, I think we should go for > the callback. Is the shift going to be handled implicitly as a > multiplicative factor handled internally by TS, or the user is > supposed to shift the output matrix? I'm inclined to implicitly > handling the shift, as it would simplify the case of constant mass > with adaptive time stepping, though it is not as general. Yes, TS should handle the shifting. > >> Why wouldn't we have this interface? >> > > As usual, the problem is on defining the interface and the trade on > performance and generality. But I'm definitely +1 on this. > >> Almost for sure MFEM must have this type of API. >> > > Stefano will love it for sure... > > > -- > Lisandro Dalcin > ============ > Research Scientist > Computer, Electrical and Mathematical Sciences & Engineering (CEMSE) > Extreme Computing Research Center (ECRC) > King Abdullah University of Science and Technology (KAUST) > http://ecrc.kaust.edu.sa/ > > 4700 King Abdullah University of Science and Technology > al-Khawarizmi Bldg (Bldg 1), Office # 0109 > Thuwal 23955-6900, Kingdom of Saudi Arabia > http://www.kaust.edu.sa > > Office Phone: +966 12 808-0459
