hyanwong / giglib

MIT License
4 stars 2 forks source link

Think about edge ordering for backward-time simulations #92

Open hyanwong opened 6 months ago

hyanwong commented 6 months ago

In backward-time simulations we will want to add older nodes at the end of the node table (allowed) but also add their edges at the end of the edge table (breaking the GIG requirements).

The algorithms will probably work fine, as long as edges are still grouped by ID. It may be that we can store edges in reverse-time and run the necessary algorithms as long as the iedge table has gigl.IEDGES_CHILD_IDS_ADJACENT even if it does not have flags = gigl.VALID_GIG (e.g. because a (not-yet-implemented) gigl.IEDGES_ORDERED_BY_CHILD_TIME flag is missing). We will probably figure out the logic when implementing #54