Save optimum variables

Specific Domains Optimization (Mathematical)
1

Dear all, I have the following optimization problem implemented:

using JuMP, CPLEX, Printf
n=50

function Problem(c)
    modelo = Model(with_optimizer(CPLEX.Optimizer))

    (n, n) = size(c)

    @variable(modelo, x[1:n,1:n], Bin)
    @objective(modelo, Min, sum(sum(c[i,j]*x[i,j] for j=1:n) for i=1:n) )
    @constraints modelo begin
        c1[i=1:n], sum(x[i,j] for j=1:n) == 10
        c2[j=1:n], sum(x[i,j] for i=1:n) == 10
    end

    status=optimize!(modelo)
    X=value.(x)
    return X
end

for i=1:10
     c=-2 .+ round.(7*rand(n,n))
     XStar = Problem(c)
end

There is a simple way to save the optimum variables “XStar” in 10 different archives (.jl or .dat) whose name depends of counter “i”?
Best regards.

2

Are you looking for binary (e.g., JLD2) or text representation (e.g., CSV)?

1 Like
3

For binary. I need to manipulate this solution in a future.

4

If “in future” you mean the same version of Julia, then:

using Serialization
X = rand(3, 3)
i = 1
open("my_file_$(i)", "w") do io
    serialize(io, X)
end

Y = open(deserialize, "my_file_$(i)", "r")

X == Y  # true

Otherwise you will need to look into JLD2 (GitHub - JuliaIO/JLD2.jl: HDF5-compatible file format in pure Julia) or similar.

1 Like