-
Notifications
You must be signed in to change notification settings - Fork 9
[MOF] Begin implementation of JuMP NLPBlock #45
Conversation
@ccoffrin the plumbing is set-up, modulo potential bugs. Thoughts on how we should test this? |
test/MOF/jump_integration.jl
Outdated
MOI.read_from_file(mof_2, "mof.json") | ||
|
||
jump_2 = Model(with_optimizer(Ipopt.Optimizer)) | ||
MOI.copy_to(jump_2, mof_2) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@mlubin, demo round-tripping a JuMP model through MOF and solving it via Ipopt.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
FWIW it's better to test the two parts separately (JuMP to MOF and MOF to JuMP) instead of testing the round trip.
YEY! Re testing, if it can write and read every model in MINLPTests.jl that's probably good, right? |
MOI.set(src, MOI.NLPBlock(), JuMP._create_nlp_block_data(src)) | ||
end | ||
MOI.copy_to(dest, backend(src); kwargs...) | ||
end |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
So the question is, should these copy_to
methods be included in JuMP? Or should they sit in MathOptFormat?
Closes #37