Ah, hm, then I am not sure what NA stands for. Even from both Readmes of the solvers, I do see that indeed they do not use derivatives, but I read NA as “Not Available” here – what does it actually stand for then?
Interestingly, I also work in non-smooth optimization, but focusing on the case where you have a subgradient or other information available (or sedulity for example).
Slight difference, but sure not the same as the X then. I was somewhere in Numerical A… as a terminology that I again do not know similar to the gradient oracle.
For anyone who is following this thread, there is a new modeling layer in Julia called ExaModels, which has shown very good performance on the AC-OPF problem, see,
You may be interested to check it out. In the same thread @sshin23 provides a nice introduction to ExaModels and the kind of use-cases where it shines.
@odow can you update the table in the first post to include an entry for ExaModels?
similar to the gradient oracle.