[Networkit] matching algorithms and data structure
henning.meyerhenke at kit.edu
Wed Oct 28 20:25:00 CET 2015
Am 28.10.15 um 18:47 schrieb Christian Staudt:
> On 28 Oct 2015, at 12:00, Moritz von Looz <moritz.looz-corswarem at kit.edu> wrote:
>> I've fixes for these issues in a private repository, can merge them into the main repo if you want to.
> Please do so comrade. You should know that privatizing bugfixes is counter-revolutionary.
>> As to b), they are useful to me right now.
> Okay. What about making Matching a subclass of Partition? I believe they are basically the same data structure.
From a conceptual point of view, this is not advisable. A matching is a
subgraph with max degree 1, whereas a Partition is, well, a partition of
the vertex set, thus something different. (A partition covers V, a
matching does not have to, for example.)
However, and that is probably what you mean, both could be implemented
in a similar way. Still, subclassing is not the way to go here. Would we
gain something by using a Partition to implement a Matching? Probably
not due to some subtle differences that would need code for special cases.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 5399 bytes
Desc: S/MIME Cryptographic Signature
More information about the NetworKit