We ignore all self-loops upon reading a particular network. Although this works fine for CPM (as self-loops have no effect there) this is not entirely correct for modularity. In particular, the problem is that the nodeWeights are calculated using getTotalEdgeWeightPerNodeHelper, but this is called only after we have removed the self-loops. In other words, the nodeWeights reflect the degree without the self-loops.
We ignore all self-loops upon reading a particular network. Although this works fine for CPM (as self-loops have no effect there) this is not entirely correct for modularity. In particular, the problem is that the
nodeWeights
are calculated usinggetTotalEdgeWeightPerNodeHelper
, but this is called only after we have removed the self-loops. In other words, thenodeWeights
reflect the degree without the self-loops.https://github.com/CWTSLeiden/networkanalysis/blob/42fdd9e4d875455b9c992501123028f5e79a713b/src/main/java/nl/cwts/networkanalysis/Network.java#L1371
In addition, we do consider self-loops when calculating the "proper" resolution parameter:
https://github.com/CWTSLeiden/networkanalysis/blob/42fdd9e4d875455b9c992501123028f5e79a713b/src/main/java/nl/cwts/networkanalysis/run/RunNetworkClustering.java#L405
Both issues should be corrected in order to make modularity work correctly when self-loops are present in a network.