Oscar Dowson
Oscar Dowson
Running @ccoffrin's (private) examples, but with https://github.com/lanl-ansi/nlp-jump-examples/blob/main/pf/nlparameter.jl updated to use the new nonlinear syntax and parameters, we now get: ```Julia julia> include("pf/base.jl"); @time bench_pf_base(case="pglib_opf_case2000_goc.m") ... 179.875041 seconds (1.05 G allocations:...
Closing because there's nothing left to do here. The key bottleneck was that JuMP rebuild the nonlinear evaluator on every call to `optimize!`. That doesn't happen anymore, so the decision...
One option from https://github.com/jump-dev/JuMP.jl/issues/3214 is to support names in containers, and then support kwarg indexing (but in the right order). something like ```Julia m = Model() @variable(m, x[item = 1:3,...
Miles says go for it
> you would have a suggestion to make a JuMP model Two options jump out: 1. a user-defined function 2. unroll the loop a fixed number of times let me...
It's not quite the same, but perhaps: ```Julia function yvec(x) y = similar(x, 8) yc = 1.6x[1] while true y[2] = yc y[3] = 1.22y[2] - x[1] y[6] = (x[2]...
Xpress has `XPRS_KEEPBASIS` or `XPRS_REFACTOR` (to fix with links)
Upstream issue is https://github.com/ERGO-Code/HiGHS/issues/1598
To clarify, what type of constraints is this iterating over?
I think we decided that this was because it queries the constraint matrix: https://github.com/jump-dev/HiGHS.jl/blob/1d192e20410b2e71c7cf1eaab48d3e533f3f8902/src/MOI_wrapper.jl#L1662-L1708 One option would be to use a cache instead of `direct_model`, so something like `model =...