R Programming/Network Analysis
We mainly use the following packages to demonstrate network analysis in R: statnet, sna, igraph. They are however not representing a complete list. See Task view of gR, graphical models in R for a complete list.
Creating simple graphs with igraphEdit
> # load the appropriate library > library(igraph) > # now create a few simple graphs > # an undirected graph with 10 nodes and without any edge > g1 <- graph.empty(10,directed=FALSE) > # a directed graph with 10 nodes > g2 <- graph.ring(10,directed=TRUE) > # a complete undirected graph with 10 nodes > g3 <- graph.full(10,directed=FALSE) > # now get information about these graphs > summary(g1) > # g1 is an igraph object, U = Undirected, with 10 nodes and 0 edge > IGRAPH U--- 10 0 -- > summary(g2) > # g1 is an igraph object, D = Directed > IGRAPH D--- 10 10 -- Ring graph
Creating graphs from dataEdit
First load the igraph package
then you can choose your preferred format. Below are examples of data provided as edge list and as adjacency matrix.
Creating graph from an edge listEdit
An edge list is formed by a two-column matrix, with each row defining one edge. An edge is drawn from each element in the first column to the corresponding element in the second one.
graph.edgelist() function to import your data.
# producing some random data in edge list form el <- cbind(sample(1:10, 10), sample(1:10, 10)) # creating and plotting the graph from the edge list gr <- graph.edgelist(el) plot(gr)
Creating graph from an adjacency matrixEdit
An adjacency matrix is a n × n matrix containing n vertices and where each entry aij represents the number of edges from vertex i to vertex j. To import your adjacency matrix, use the
# producing a random adjacency matrix adj <- matrix(sample(0:1, 100, replace=T), 10, 10) # creating and plottig the graph from the adjacency matrix gr <- graph.adjacency(adj) plot(gr)
- Statnet website includes all the documentation on network analysis using R.
- Julien Barnier's introduction (in French)
- Journal of Statistical Software #24 Special Issue on Networks in R