Oscar Dowson

Is the claim something like this: ```Julia julia> import MathOptInterface as MOI julia> x = MOI.VariableIndex.(1:2) 2-element Vector{MathOptInterface.VariableIndex}: MOI.VariableIndex(1) MOI.VariableIndex(2) julia> model = MOI.Nonlinear.Model(); julia> objective = :(sin($(x[1])) + $(x[2])^1)...

Your `@NLconstraints` are actually quadratic, so you don't need to do AD on them. Here's the underlying issue: ```Julia julia> x = MOI.VariableIndex.(1:2) 2-element Vector{MathOptInterface.VariableIndex}: MOI.VariableIndex(1) MOI.VariableIndex(2) julia> model =...

So I assume somewhere we are detecting Hessian sparsity and if `op(x, y)` then we are adding `(x, x), (x, y), and (y, y)`. We should special-case `*`.

I think this is even somewhat expected from our sparsity algorithm. We don't recognize that `sum(x) / y` is decomposable. The sparsity algorithm trades simplicity for some false positives. I...

Perhaps this issue is really asking for better insight on how different formulations might perform: ```Julia julia> using NLPModels, NLPModelsJuMP, JuMP, SparseArrays julia> function nlp_hessian(model) nlp = MathOptNLPModel(model) rows, cols...

This came up again in https://discourse.julialang.org/t/too-many-nonzeroes-for-nlp-in-jump-and-examodels/123901

I think I'm going to close this issue. There's certainly scope for writing a new AD backend, but I'm very reticent to go making changes to our existing one. The...

Potentially related to https://support.gurobi.com/hc/en-us/requests/66420 I've been trying to track down a segfault for dkee331. I don't think I can share the model because it contains some industry data. ``` julia>...

Closed by #552

> So, almost 2/3 of the time is spent on adding the bounds. I would expect this, since there are three things we need to do: add a variable, add...