Preconditioner about argmin HOT 1 OPEN

Preconditioning would be a nice feature! If I understand it correctly, it only requires another multiplication with a matrix (or the inverse of a matrix) with a vector. I'm however unsure how this is best done in practice. I've used preconditioning by applying the preconditioning to the system itself, but not as part of the solver. I guess that, depending on the type of preconditioning, there may be computationally efficient implementations, and we should not limit the implementation such that these become impossible. But I don't have a sufficient overview over the topic. Do you by any chance know how this is solved in other solvers?

Could you elaborate how you would solve this with a trait? From the top of my head I'd say that an optional field holding the preconditioner may be sufficient, at least for a basic implementation.

Is there a minimum-effort way to compare our algo? Like, using what you already coded to update the array or some magic that you know?

Do you want to compare performance (speed) or performance (convergence, ...)? For the latter I suppose choosing your preconditioner such that it has no effect would work. The former probably requires some code.

from argmin.