codeGCluster: codeGCluster: A Graph Clustering Algorithm Based on Community...

In XiaoyuLiang/i2d: Convert image to digital data

Description

The community detection-based graph clustering algorithm (codeGCluster) is baed on community detection methods, such as Louvain algorithm, fast-greedy modularity optimization algorithm, and near linear time algorithm. codeGCluster retruns community information.

Usage

codeGCluster(dat=dat, wt=4, k=NULL, diff_distance=0.1, method="louvain");

Arguments

dat	data.frame or matrix of x-coordinate and y-coordinate.
wt	weight: the initial distance cut-off value. The weight of edge larger than wt will be removed.
k	integer: number of clusters. If k=NULL (the default), graph will be automatically clustered using a community detection method in igraph package. Otherwise, graph can be forced to cluster into k clusters.
diff_distance	numeric value: distance change for each iteration.
method	character: method for community detection. The method should be one of "louvain", "fast_greedy", "infomap", "label_prop", or "spinglass".

Author(s)

Ying Hu [e-mail]

Chunhua Yan [e-mail]

Xiaoyu Liang [e-mail]

References

Csardi, G., & Nepusz, T. (2006). The igraph software package for complex network research. InterJournal, complex systems, 1695(5), 1-9.

Blondel, V. D., Guillaume, J.-L., Lambiotte, R., & Lefebvre, E. (2008). Fast unfolding of communities in large networks. Journal of statistical mechanics: theory and experiment, 2008(10), P10008.

Clauset, A., Newman, M. E. J., & Moore, C. (2004). Finding community structure in very large networks. Physical review E, 70(6), 066111.

Raghavan, U. N., Albert, R., & Kumara, S. (2007). Near linear time algorithm to detect community structures in large-scale networks. Physical review E, 76(3), 036106. Reichardt, J., & Bornholdt, S. (2006). Statistical mechanics of community detection. Physical review E, 74(1), 016110.

Examples

data(Cluster_Example_3,package="i2d")
set.seed(1234)
dat <- i2d(image=Cluster_Example_3, p.n=500)

## Perform graph clustering method based on a community detection algorithm
out <-codeGCluster(dat=dat, wt=4, k=4, method="louvain",diff_distance=0.5)

## Plot the clutered vertices
cols <- rainbow(12, alpha=0.5)
plot(dat, pch=19, col=cols[out$membership], cex=1.2)

XiaoyuLiang/i2d documentation built on Oct. 20, 2020, 11:05 p.m.