Thanks @Gary!
@Antti-L, please try this one: https://github.com/olejandro/TIMES.jl/tr...d3d8a479a5
I've changed YEAR to MODLYR in the definition of the variables to avoid generating unnecessary variables. It is interesting to see that this wasn't necessary in the Pyomo example.
On my laptop, the Utopia model now takes about 1 minute, which is still much slower than the GAMS and Pyomo examples. But I guess, it should now be okay to try it on your Proto model?
I'll look into other common issues (e.g. type instability) that influence Julia / JuMP performance.
@Antti-L, please try this one: https://github.com/olejandro/TIMES.jl/tr...d3d8a479a5
I've changed YEAR to MODLYR in the definition of the variables to avoid generating unnecessary variables. It is interesting to see that this wasn't necessary in the Pyomo example.
On my laptop, the Utopia model now takes about 1 minute, which is still much slower than the GAMS and Pyomo examples. But I guess, it should now be okay to try it on your Proto model?
I'll look into other common issues (e.g. type instability) that influence Julia / JuMP performance.