Joshua Pulsipher

Some plans on this were shared at https://www.youtube.com/watch?v=4DwpaqxAKwU

The update to BARON.jl might be a helpful guide: https://github.com/jump-dev/BARON.jl/pull/63

@hdavid16 here is a simple illustrative use case: ```math \begin{gathered} \left[\begin{gathered} Y_i(\xi)\\ x_i(\xi) \leq 0 \end{gathered}\right] \vee \left[\begin{gathered} \neg Y_i(\xi)\\ x_i(\xi) \geq 0 \end{gathered}\right], \ \ i \in I, \xi...

> @pulsipher this is an interesting application because you are suggesting using cardinality constraints in a conditional fashion. The current modeling paradigm in GDP is that cardinality constraints are global...

Expanding on the original issue, some reformulations that would be nice to implement include: - [ ] Improved big-M (https://www.sciencedirect.com/science/article/pii/S0098135415000587) - [ ] P-Split (https://arxiv.org/abs/2202.05198) - [ ] Cutting-plane (https://doi.org/10.1016/S1570-7946(03)80444-3)...

As an update, it seems MOI bridges will likely not work with disjunctions. The problem is that each disjunct can contain an arbitrary number of constraints of any type (including...

After talking offline with @hdavid16, another option to is add a canonicalized disjunction set where all the constraints are converted to inequalities first: ```julia struct DisjunctionSet{T

@bernalde and @pedromxavier I think this PR gives you the MOI representation that you need. Namely, the addition of `DisjunctionSet

> Should there be a `MOI.get` interface for accessing the contents of the Disjunction Set? Any suggestions on what the input and output would be precisely?

@hdavid16 I think it may be worth keeping this feature request open to only use one logical variable when there are two disjuncts. For linear/quadratic problems that are compatible with...