Extracting solution of decision variables

[binkob]

@sshin23

When solving the examodel, the result is set up in the form:

result = madnlp(em),

however, I could not find a way to extract the exact solution for a variable

I would like to know how specific variables can be extracted. I realized that using ".solution" gets the solution of all variables, but they are combined. For large scale, that can be a problem. I could not find it in the examples.

Ex: If I define variables p and q as decision variables, is there a way I can extract its solution only?

sol_p = value.(p) as used in JuMP.

Thanks

[sshin23]

Hi @binkob, have you checked the link below?

https://exanauts.github.io/ExaModels.jl/dev/guide/

[binkob]

@sshin23
The example worked with the direct examodel formulation. However, I have my formulation in JuMP and used the JuMP->Examodel conversion.
I know it's in the experimental stage, but I found it flexible for my application.

ERROR: type Array has no field offset

I get this error when after solving and I try to fetch variable x.

I also run the jump.jl example and got the same error.

I might be missing something.

using ExaModels, JuMP, CUDA

N = 10
jm = Model()

@variable(jm, x[i = 1:N], start = mod(i, 2) == 1 ? -1.2 : 1.0)
@constraint(
jm,
s[i = 1:N-2],
3x[i+1]^3 + 2x[i+2] - 5 + sin(x[i+1] - x[i+2])sin(x[i+1] + x[i+2]) + 4x[i+1] -
x[i]exp(x[i] - x[i+1]) - 3 == 0.0
)
@objective(jm, Min, sum(100(x[i-1]^2 - x[i])^2 + (x[i-1] - 1)^2 for i = 2:N))

em = ExaModel(jm; backend = CUDABackend())

em = ExaModel(jm) #; backend = nothing)

Here, note that only scalar objective/constraints created via @constraint and @objective API are supported. Older syntax like @NLconstraint and @NLobjective are not supported.

We can solve the model using any of the solvers supported by ExaModels. For example, we can use MadNLP:

using MadNLPGPU

result = madnlp(em)

x = solution(result,x)

Thank you for your time.

[sshin23]

If you are using JuMP interface and want to access the solution, you may use the following interface:

using ExaModels, JuMP, CUDA
using MadNLPGPU

set_optimizer(jm,  ExaModels.MadNLPOptimizer)
optimize!(jm)