Closed blegat closed 10 months ago
This is why it was not detected in the tests: https://github.com/JuliaSmoothOptimizers/NLPModelsJuMP.jl/pull/167#issuecomment-1764087888
Ok... I am slighly surprised it ever worked now as I don't remember the last we updated the GenericExecutionStats
constructor. The main one being:
stats = GenericExecutionStats(nlp)
because it allocates some structure that are dependent of the size of the problem.
The solution might be to proceed in the same way we handle the solver?
It should be possible to make an empty constructor for the GenericExecutionStats
, but it will at minima requires knowing the type used for the solution vector and the multipliers (for instance, Vector{Float64}
, etc...).
Ok thanks, https://github.com/JuliaSmoothOptimizers/NLPModelsJuMP.jl/pull/168 should fix it by simply using nothing
.
Do you want a new release?
Yes please :)
Ok, I will do it later today :+1:
@tmigot Do you plan to open a PR with your new empty GenericStatsExectution that you added in SolverCore.jl
?
Thanks @amontoison , I will add it later, no rush there.
It seems the constructor
NLPModelsJuMP.Optimizer()
broke with an update of one of the dependencies. When running the tests, I getThe versions used are