-
Notifications
You must be signed in to change notification settings - Fork 123
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Tidy src/variable.jl #583
Tidy src/variable.jl #583
Conversation
Yeah, that was the plan. After using it intermittently though I don’t really like the *Var suffixes; they are hard to remember and look a bit weird. I’m definitely open to alternative approaches! |
What about |
Either of those seems better! The MOI ones might help folks connect the dots with MOI more, while the Base ones might look less scary. Both could lead to the issue of folks trying other objects that aren’t supported, but we could give good errors. I see JuMP supports |
Btw now that JuMP supports generic number types, it seems ok to lower to jump rather than MOI- maybe that could allow better interop, like mixing Convex and JuMP variables in a model. I bring this up here since I suspect |
I think we want to keep lowering to MOI instead of JuMP. We have some ideas for DCP in JuMP. One is to lower JuMP's nonlinear expressions to Convex (requires jump-dev/MathOptInterface.jl#2402). The other is to add DCP support to JuMP directly. Working through this code base is giving me some ideas. But this is a long-term thing. Nothing in the short-term. |
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## master #583 +/- ##
==========================================
+ Coverage 97.76% 97.94% +0.18%
==========================================
Files 91 91
Lines 5060 5069 +9
==========================================
+ Hits 4947 4965 +18
+ Misses 113 104 -9 ☔ View full report in Codecov by Sentry. |
You can already mix any constraints supported by JuMP with DCP constraints when using |
|
Ah, but it is the symbol
Perhaps this is a separate PR through. |
@ericphanson thoughts on
:Bin
vsBinVar
? I assume your long-term plan was to deprecate the symbols? They're pretty brittle and the docs don't mention the Symbol approach.