Jutho
Jutho
Note, by the nature of the `if` clause, that you will not run into this problem if you have a single threaded Julia instance. But that may not be an...
Thanks for the suggestion. That would not be entirely straightforward to implement, as that keyword argument would need to penetrate from the high-level methods such as `eigsolve` and `linsolve` to...
Cool; I'll try to review as soon as possible.
Hi Miles, thanks for the report. I will take a look. Certainly this is very different then how `tol` works for linear problems and eigenvalues, as for `exponentiate` there is...
Actually, it turns out the `eager` mode was already implemented. My memory is terrible for these things. So just add `eager=true` to your keywords and you will see the desired...
The most recent commit adds documentation for the `eager` behaviour in `exponentiate` and `expintegrator`. I will close this issue; feel free to reopen if you think anything else needs to...
I've reopened this for further discussion. Regarding the tolerance, also note that (as mentioned in the docs) `tol` is the requested accuracy per unit time. This means that, in your...
KrylovKit still uses the julia LinearAlgebra standard library, or thus BLAS and LAPACK, for dealing with the small problems projected on the Krylov subspace. So it is currently restricted to...
It could, but so far I did not find the time to implement these backends. It also seems there is some fragmentation of linear algebra routines for generic numbers /...
Indeed, I assume this should have been changed a long time ago already. Thanks