Converting between graph classes

As mentioned above, a common network analysis workflow includes the conversion of graph objects between various formats. The igraph and statnet packages provide some very basic functions for this, that are specific to the types of objects that are used in these packages themselves:

Convert to an adjacency matrix
Graph object conversion
igraph	`igraph::as_adjacency_matrix(g, sparse = FALSE)`
network	`network::as.sociomatrix(g) network::as.matrix.network(g)`
Convert to an adjacency list
igraph	`igraph::as_adj_list(g)`
network	—
Convert to an edge list
igraph	`igraph::as_edgelist(g) igraph::as_data_frame(g)`
network	`network::as.data.frame.network(g) network::as.edgelist(g) # if you want to include edge weights sna::as.edgelist.sna(g)`
Make the network directed
snafun	—
igraph	`igraph::as.directed(g)`
network	—
Make the network undirected
snafun	`# from an adjacency matrix as input snafun::to_symmetric_matrix(g, rule = c("weak", "mutual", "out", "in", "average", "max", "min"), na.rm = TRUE)`
igraph	`igraph::as.undirected(g) # for example igraph::as.undirected(g, mode = 'collapse', edge.attrib.comb = list(weight = 'sum'))`
network	`# from an adjacency matrix as input sna::symmetrize(g) # for example sna::symmetrize(g, rule = "strong")`
Project a bipartite graph
snafun	—
igraph	`igraph::bipartite_projection(g)`
network	—
Convert to a line graph
snafun	—
igraph	`igraph::make_line_graph(g)`
network	—

Conversion between various formats is a lot easier with the help of the snafun package. You only need to know some very easy function names to convert between the common graph classes. Here is an overview of which function to use for which conversion:

INPUT	OUTPUT
Convert between various formats
Using the consistent functions of snafun
INPUT	edgelist	matrix	igraph	network
edgelist	`to_edgelist()`	`to_matrix()`	`to_igraph()`	`to_network()`
matrix	`to_edgelist()`	`to_matrix()`	`to_igraph()`	`to_network()`
igraph	`to_edgelist()`	`to_matrix()`	`to_igraph()`	`to_network()`
network	`to_edgelist()`	`to_matrix()`	`to_igraph()`	`to_network()`
Conversion includes bipartite graphs. Check the arguments for the various options.