Thomas Baumann

Are you sure the implementation details are set in stone at this point? You can consider adding the tutorial as documentation to the actual code of the class you implement...

Don't worry, it's not difficult. I did it recently in #417. You can use `.. literalinclude::` to include a file of code or any txt file or whatever into the...

Oh don't worry! I am quite eager to work on this, but it's gonna be a whole thing! I realised that I was constructing linear multistep methods as preconditioners and...

So apparently there is a theorem that there are no A-stable LMMs with order higher than two, which makes these preconditioners somewhat useless... But they are convergent at least.

The first order LMM that is derived in this way is backward Euler and the second order LMM that is derived here is the trapezoidal rule, both of which are...

We don't have to merge this in the master branch by the way. I will maybe look more into the regions of stability (which depend on the nodes that you...

My plan was to make script that executed dahlquist problems for some patch in the complex plane and shows if the scheme is stable. Do you think it's possible to...

Ok that was what I was about to do, so that's good! The many-tests-simulateneously Dahlquist equation is already implemented in pySDC. I saw that the stability region depends on the...

The plots are here: https://github.com/Parallel-in-Time/pySDC/blob/35535b6b1bcdd4038d77047897f7ea8ff5b2f682/pySDC/playgrounds/Preconditioners/LMM_preconditioner.ipynb Doesn't look too good...

> > So apparently there is a theorem that there are no A-stable LMMs with order higher than two, which makes these preconditioners somewhat useless... But they are convergent at...