Oscar Dowson
Oscar Dowson
We already update here: https://github.com/jump-dev/Gurobi.jl/blob/4d204e22cf0f6004e9580583480835c451ebbf97/src/MOI_wrapper/MOI_wrapper.jl#L3691-L3698 So is the only thing missing an update at the start of `set`? https://github.com/jump-dev/Gurobi.jl/blob/4d204e22cf0f6004e9580583480835c451ebbf97/src/MOI_wrapper/MOI_wrapper.jl#L3734-L3752
I'm okay if we throw the same warnings as Python. But perhaps you could fix this internally to only throw a warning if `GRBoptimize` is called without a complete basis?
Okay. Does this solve #516? Or should we still buffer the deletions?
Ugh. Tests pass locally on `nightly`. But not on CI...
I think there's something deeper at play. Hard to replicate because I don't see these numbers locally?
Yeah can't you use an intermediate callback to store this information?
Wait up, why is this method even here??? 🏴☠️ https://github.com/lanl-ansi/PowerModels.jl/blob/a8d1aadeb0993b4efb706f4197e4f6178354a3db/src/core/objective.jl#L7
But yes, there are probably a bunch of places that we need to replace hard-coded constants with the appropriate `T(x)` version. I don't know if there's anything fundamental limiting, it's...
@blegat is this expected? I think it's because you don't actually have a free decision variable in your model?
Your first example is expected behavior for broadcasting. ```Julia julia> x = rand(3, 4) 3×4 Matrix{Float64}: 0.724199 0.0505117 0.789073 0.977107 0.203153 0.245935 0.651512 0.641174 0.0745526 0.138428 0.0774511 0.900541 julia> sum(x;...