Instantiation of `NoCriterion` fails about benchopt

[maltezfaria]

This was on Version: 1.4.0 if that helps.

[mathurinm]

Not sure, but this may be legacy code : it's not used in the benchopt repo.

What do you mean by, "I want to control the stopping criterion manually, using max_iter" ? isn't sampling_strategy = "iteration" enough ? if you want a finer control on the sequence of max_iter used by your solver, you can control it via the get_next method (see en of this paragraph : https://benchopt.github.io/performance_curves.html?highlight=how%20sampling%20curves#using-iterations-or-tolerance)

Or do you want a single run ? In the case SingleRunCriterion would be the way to go.

[maltezfaria]

You are right, I don't really need it: get_next does what I want. Thanks!

[maltezfaria]

This bug is related to legacy functionality no longer used in benchopt, closing the issue.