Adding nodes to a subgraph after the main graph's backend is initialized does not reset graph backend

[dlcole3]

I have found that adding to a subgraph does not necessarily update the overall graph's backend. For example, the code below throws an error because the first incident_edges call sets the backend of the graph, and then when the second node is added, the backend on graph g0 is not updated. When Plasmo runs the second call of incident_edges, it does not have the extra_node in its hypermap.

using Plasmo

g0 = OptiGraph()
g1 = OptiGraph()
g2 = OptiGraph()

@optinode(g1, node)
@variable(node, x >= 0)

@optinode(g2, node)
@variable(node, x >= 0)

add_subgraph!(g0, g1)
add_subgraph!(g0, g2)

incident_edges(g0, all_nodes(g1))

@optinode(g1, extra_node)
@variable(extra_node, x >= -5)

incident_edges(g0, all_nodes(g1))

This does make sense since there is no mapping to go from the subgraph to its "owning" graph (only to go from the "owning" graph to the subgraph), so there is no way for the upper graph to know to update its backend. Currently, I get around this in my code by forcing the upper graph to reset the backend by calling Plasmo.set_graph_backend, but if there is an efficient way to fix this in the future, it could be nice.

[jalving]

This may be a good reason to maintain a reference from optinodes to each of their containing optigraphs. The original design of Plasmo.jl actually did this, but we opted to make nodes more modular at the time. This will be added to the roadmap in the re-write.

[jalving]

This is still an issue in v0.6. The main problem is that the projection used to query graph topology is not kept in sync with the optigraph. In this case, you need to create a new projection after you update the optigraph. Here is a working example:

using Plasmo

# setup example optigraph
g0 = OptiGraph()
g1 = OptiGraph()
g2 = OptiGraph()
@optinode(g1, nodes1[1:2])
@variable(nodes1[1], x >= 0)
@variable(nodes1[2], x >= 0)
add_edge(g1, nodes1[1], nodes1[2])
@optinode(g2, nodes2[1:2])
@variable(nodes2[1], x >= 0)
@variable(nodes2[2], x >= 0)
add_edge(g2, nodes2[1], nodes2[2])
add_subgraph(g0, g1)
add_subgraph(g0, g2)
add_edge(g0, nodes1[1], nodes2[1])

# create projection and query incident edges
projection = hyper_projection(g0)
incident_edges(projection, all_nodes(g1))

# update optigraph
@optinode(g1, extra_node)
@variable(extra_node, x >= -5)
add_edge(g1, extra_node, nodes1[1])

# this will fail because the projection is not kept in sync
incident_edges(projection, all_nodes(g1))

# create a new projection instead
new_projection = hyper_projection(g0)
incident_edges(new_projection, all_nodes(g1))

At some point, we should consider developing methods to update the projection in cases where a user needs to modify the projection extensively.