Add IMEX time-stepper interface and Heun2 IMEX #5617
Conversation
wthrowe
force-pushed
the
imex/ImexTimeStepper
branch
from
bade47c
to
8bb706e
Compare
| * \f$Y_{n,\text{inhomogeneous}}\f$ is the contribution of the past | ||
| * values from the implicit derivative history, and \f$S(\cdot)\f$ is | ||
| * the implicit portion of the right-hand-side (generally source | ||
| * terms for PDEs). We call \f$w_n\f$ the \emph{implicit weight}. |
There was a problem hiding this comment.
Doxygen doesn't like \emph. Need to replace it with something else. (Not sure how I messed that locally).
wthrowe
force-pushed
the
imex/ImexTimeStepper
branch
from
8bb706e
to
efb56ad
Compare
| namespace TimeSteppers { | ||
| /*! | ||
| * \ingroup TimeSteppersGroup | ||
| * Intermediate base class implementing a generic IMEX Runge-Kutta |
There was a problem hiding this comment.
[optional] One of the best discussions (not to say I find it crystal clear) of IM(EX) RK methods I came across is this one from NASA: https://ntrs.nasa.gov/api/citations/20160005923/downloads/20160005923.pdf Might be worth linking/referencing if you think it could be useful to people? Numerical Recipes doesn't really discuss this in a lot of detail so having something else for folks to learn about it with could be good. Maybe also reference the second Hairer+ book?
There was a problem hiding this comment.
I added the NASA reference here, since it explains the notation well. I'll include Hairer in the tutorial, since that has a bit of discussion of properties of implicit methods and Hairer covers that in depth.
| struct ImplicitButcherTableau { | ||
| /*! | ||
| * The coefficient matrix of the substeps. We can only reasonably | ||
| * support EDIRK methods, so the tableau must be lower-triangular |
There was a problem hiding this comment.
What about ESDIRK and QESDIRK? I don't know the detailed differences, so I would find it helpful to list those explicitly as either being okay or not. I think this rules out collocation-type methods, at least as the most obvious ones to me. Is that right? Might also be worth stating explicitly
wthrowe
force-pushed
the
imex/ImexTimeStepper
branch
from
efb56ad
to
2472e33
Compare
wthrowe
force-pushed
the
imex/ImexTimeStepper
branch
from
2472e33
to
35c6311
Compare
Proposed changes
Upgrade instructions
Code review checklist
make docto generate the documentation locally intoBUILD_DIR/docs/html.Then open
index.html.code review guide.
bugfixornew featureif appropriate.Further comments