Hello @jakobsj and @jeverink. I now refactored and fixed conflicts with main. I also updated the regression tests that needed to be updated after the bug fix to FISTA. As for jakobs point 4)
"I think the proximals should be put somewhere dedicated, not directly in cuqi.solver, as they are not solvers. Could be something like cuqi.solver.proximal or rename as cuqi.opti.proximal, and could consider moving e.g. FISTA, PDHG etc. to cuqi.opti.algo or similar, since "solver" is a bit vague, could mean equation solver or other solver, and I think all are optimization algorithms."
I would like to postpone fixing this and creating an issue to fix it. It would be a rather substantial change that would be best handled in a new PR. I would also like to avoid creating a new temporary module (e.g. cuqi.proximals) which we may delete after refactoring cuqi.solver.
"I think the proximals should be put somewhere dedicated, not directly in cuqi.solver, as they are not solvers. Could be something like cuqi.solver.proximal or rename as cuqi.opti.proximal, and could consider moving e.g. FISTA, PDHG etc. to cuqi.opti.algo or similar, since "solver" is a bit vague, could mean equation solver or other solver, and I think all are optimization algorithms."
I would like to postpone fixing this and creating an issue to fix it. It would be a rather substantial change that would be best handled in a new PR. I would also like to avoid creating a new temporary module (e.g. cuqi.proximals) which we may delete after refactoring cuqi.solver.
Originally posted by @nabriis in https://github.com/CUQI-DTU/CUQIpy/issues/291#issuecomment-1894531646