I apologize if someone had raised questions like this. And I understand the topic sounded similar to some previous topics. I did do a search. But didn’t found specific answers to all the questions I have. I understand that Gurobi is not a nonlinear programming solver. But it’s one of the lead s…

[image] kguo: Is it worth to go through all the trouble trying to reformulate the problem somehow into a convex quadratic problem approximately or a nonlinear solver such as Ipopt works as well? I don’t think there’s any general consensus on this. Best is to try and see. [image] kguo: Is…

These are very helpful. I think that you meant a different order on the second and third augment, like JuMP.register(model, :outP1, 4, (x1,x2,x3,x4) -> outP(i,x1,x2,x3,x4), autodiff=true) And that worked.

[image] kguo: What if x is of high dimension? Any way to get around the augment restriction or some other clever ways to handle nonlinear constraints? If you use e.g. NLopt directly, you can employ high-dimensional (e.g. I’ve used 10⁵-dimensional x) nonlinear constraints with arbitrary user-…

I think I understand registration and the need for closure. Since we need to iterate (the need for outP1, outP2, …), we need to use symbol. We’ve managed to have that work. However, splatting seems to be a problem. It didn’t work together with symbol. Here is the block of code. for i=1:K udfS…

I’ve used this approach before . Edit: which I see now is basically just what Miles proposed earlier.

Sorry, it seems that I only come back to look at this problem once every month this summer. But could someone point out to me what I did wrong with this simple piece of code? arg_expressions = [:($(x[i])) for i=1:4] JuMP.register(model, :outP, 4, outP, autodiff=true) #@NLobjective(model, Min, ou…

I’m really weak with expressions and nonlinear modeling in JuMP, so hopefully @miles.lubin can explain the right way to do this. In the mean time, you can try building the expression yourself and using JuMP.setNLobjective like so: julia> function my_obj(x1, x2, x3, x4) return (x1-1)^2 +…

Can you please tell me how did you install IPOPT solver? I am getting build issues with it. I do have the solver binaries. How do I use them and can that be generalized for other solver binaries too? I am using JuMP in JuliaPro. Thanks

The Ipopt binaries are provided on OS X which is what I’m using. I assume that you are using Linux which will be compiled from source. Sorry, I don’t have experience with that. If you post it to the mailing list, I’m sure someone will have experience with it.

Solvers, nonlinear constraints and user defined functions

Specific Domains Optimization (Mathematical)

kguo September 15, 2017, 2:31pm 9

Thanks! This helps a lot. I think your explanation of its being one argument with type expression makes sense. I had similar problem with @NLconstraint as well. I think it’s for the same reason.

Topic		Replies	Views
How to add user-defined nonlinear functions as constraints with JuMP? Optimization (Mathematical) jump	3	2100	June 5, 2018
Adding nonlinear constraints in JuMP without hard-coding Optimization (Mathematical)	4	1104	January 18, 2022
Variable number of Nonlinear Constraints by user-defined functions Optimization (Mathematical) question	1	865	September 19, 2017
JuMP error for nonlinear optimization Optimization (Mathematical) jump , nonlinear	17	1988	November 6, 2021
Constraints of type MathOptInterface.ScalarNonlinearFunction-in-MathOptInterface.LessThan{Float64} are not supported by the solver Optimization (Mathematical) nonlinear , gurobi , nonlinear-optimizati , nonlinear-programmin	7	172	February 23, 2025

Solvers, nonlinear constraints and user defined functions

Related topics