Skip to content

Moving pgr_sequentialVertexColoring to proposed on 3.3 #2202

New issue

Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? # to your account

Closed
krashish8 opened this issue Oct 25, 2021 · 0 comments
Closed

Moving pgr_sequentialVertexColoring to proposed on 3.3 #2202

krashish8 opened this issue Oct 25, 2021 · 0 comments

Comments

@krashish8
Copy link
Member

The function pgr_sequentialVertexColoring - Experimental under the traversal family of functions can be moved to proposed on version 3.3. Function's documentation can be found on https://docs.pgrouting.org/latest/en/pgr_sequentialVertexColoring.html

This function has all the types of tests that are required, and hence is less likely to crash the server. It has a single signature and is based on the straightforward greedy graph coloring algorithm. The tests that are present:

  • No crash tests, when the parameters are NULL.
  • Types check
  • Inner queries check
  • Edge cases on a subset of edges & vertices of the Sample Data (18 tests)

The function belongs to a new family of function (Coloring family), and can be used to color a graph such that no adjacent vertices have the same color. In other words, it can be used as an alternative to (not-so-optimal) job scheduling such that jobs (= vertices) with clashing constraints are colored differently (= assigned different slots). It can be also used as a way of checking whether a graph is bipartite (i.e. when only two colors can be used to color a graph), or dividing some set of clashing tasks into two sets of non-clashing tasks.

Hence, this function can be moved to proposed on 3.3.

cc: @cvvergara

# for free to join this conversation on GitHub. Already have an account? # to comment
Projects
None yet
Development

No branches or pull requests

2 participants