Oscar Dowson
Oscar Dowson
> This results in twice as many cuts being present at node 1 than other nodes I think that makes sense. You need a cut for `x0` and a cut...
Yeah I'm not sure. I have't played with this much so something probably needs fixing. Perhaps we should return `scenario_path [1:end-1]` here https://github.com/odow/SDDP.jl/blob/5687376e0ed92009a42e2c140edb24bb0c371740/src/plugins/sampling_schemes.jl#L272
And if you don't use `@stageobjecive` you get a `+(Nothing, AffExpr)` error.
Ah nice! I've been trying to find a test-case for this: https://github.com/jump-dev/Gurobi.jl/issues/415. Can you run it to get the Gurobi log? You can save the file with `SDDP.write_to_file(model, "gurobi_failure.mof.json.gz"; test_scenarios...
> It seems for the constraints are changed and all scenarios need to follow those constraints. What I would like to achieve is to have scenario specific constraints. According to...
> it gives me this error: Yeah you need to call `write_to_file` before `train`. I'll take a look at the quadratic error. Do you have the full stack trace? >...
@haoxiangyang89 how do I reproduce this? I tried running the `ms_sddp.jl` file, and I got `ci not defined`.
I tried with `ci=1` and the constraints you added removed, but it runs okay. Also: what is going on with your bound? It's not monotonic. Try with `cut_deletion_minimum = 10_000`.
Moving #499 here
Found this gem of a stopping rule in MSPPy (#208) and it's somewhat related