[Networkit] Request for Comments: Efficient Set Operations on Graphs

Christian Staudt christian.staudt at kit.edu
Wed Nov 4 21:56:42 CET 2015


I’ve started the branch “setops” for any experiments in this direction. Feel free to try out some code (but not in the Dev branch due to the upcoming release next week).


On 04 Nov 2015, at 12:29, Christian Staudt <christian.staudt at kit.edu> wrote:

> 
> On 04 Nov 2015, at 11:32, Matteo Riondato <matteo at cs.brown.edu> wrote:
> 
>> One additional comment: are these operations modifying one of the two Graph object or are they creating a new Graph (I vote for this second option, new Graph). I’m asking because while union() and intersect() should create a new graph (in the same way that the set operation “create" a “new” set), append() is more ambiguous, and actually suggests that one of the Graph object is modified, but I would suggest some uniformity among these (or non-uniformity, my point is that it must be a motivated design choice).
> 
> 
> That’s a good question. I think they should be modifiers on an existing graph, e.g. void Graph.append(Graph).  For my application, I want to extend a possibly already large graph G by adding small subgraphs. In this case, I do not want a copy of G enforced. For uniformity reasons, union and intersect should be modifiers too, right? If you want a new graph, you can always make a copy before applying the operation.
> 
> 
> 
> _______________________________________________
> NetworKit mailing list
> NetworKit at ira.uni-karlsruhe.de
> https://lists.ira.uni-karlsruhe.de/mailman/listinfo/networkit

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 495 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <https://lists.ira.uni-karlsruhe.de/mailman/private/networkit/attachments/20151104/61e31b75/attachment-0001.sig>


More information about the NetworKit mailing list