Difference between revisions of "Social Network Analysis With Igraph Package Using R"
From AIRWiki
(→Starting with the package) |
|||
| (16 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
| − | = Install the package under R environmemnt = | + | == Install the package under R environmemnt == |
| − | + | under Ubuntu install the R package: | |
| − | <code>install | + | <code>sudo apt-get install r-base-core</code> |
| + | type R to start the program: | ||
| + | <code>R</code> | ||
| − | = Starting with the package = | + | type the command to download and install igraph as R library: |
| + | <code>install.packages("igraph")</code> | ||
| + | |||
| + | == Starting with the package == | ||
don't forget to load the library: | don't forget to load the library: | ||
<code>library('igraph')</code> | <code>library('igraph')</code> | ||
parse a network file in Pajek format and access to its memory representation through the variable 'g': | parse a network file in Pajek format and access to its memory representation through the variable 'g': | ||
| − | <code>g <- read.graph(file='/your/path/.../file.net', format='pajek')</code> | + | <code>g <- read.graph(file='/your/path/.../file.net', [[pajek graph data format|format='pajek']])</code> |
list vertex and their attributes of graph g: | list vertex and their attributes of graph g: | ||
<code> | <code> | ||
| − | V(g) # the list of all nodes | + | V(g) # the list of all nodes |
| − | V(g)[0] # access the first node | + | V(g)[0] # access the first node |
list.vertex.attributes(g) # the list of all vertex attributes names | list.vertex.attributes(g) # the list of all vertex attributes names | ||
| − | V(g)[0]$id # access the attribute 'id' for the first node | + | V(g)[0]$id # access the attribute 'id' for the first node |
</code> | </code> | ||
| + | list edges/arcs and their attributes of graph g: | ||
| + | <code> | ||
| + | E(g) # the list of all edges/arcs | ||
| + | E(g)[0] # access the first edge/arc | ||
| + | list.edge.attributes(g) # the list of all edges/arcs attributes names | ||
| + | E(g)[0]$weight # access the attribute 'weight' for the first node | ||
| + | </code> | ||
| + | draw the graph: | ||
| + | <code> | ||
| + | plot(g) | ||
| + | tkplot(g) | ||
| + | </code> | ||
| + | |||
| + | == Macroscopic Analysis == | ||
| + | '''graph size''' (number of nodes): length(V(g)) | ||
| + | |||
| + | '''number of edges''': length(E(g)) | ||
| + | |||
| + | '''diameter''': diameter(g) | ||
| + | |||
| + | '''mean degree''': | ||
| + | |||
| + | '''density''': | ||
| + | |||
| + | '''giant component size''' (the largest connected component size): | ||
| + | |||
| + | assortativity: | ||
| + | |||
| + | == Node Centrality Analysis == | ||
| + | |||
| + | '''degree''': degree(g) | ||
| + | |||
| + | '''closeness''': closeness(g) | ||
| + | |||
| + | '''betweenness''': betweenness(g) | ||
| + | |||
| + | '''eigenvector'''/'''pagerank''': evcent(g) | ||
| + | |||
| + | == Clustering Techniques == | ||
| + | |||
| + | '''connected components''': clusters(g) | ||
| + | |||
| + | '''edge betweenness''': | ||
| + | |||
| + | '''walktrap''': | ||
| + | |||
| + | '''spinglass''': | ||
| + | == References and further informations == | ||
| + | Igraph/R documentation: http://igraph.sourceforge.net/doc/R/00Index.html | ||
[[Category:Tutorial]] | [[Category:Tutorial]] | ||
Latest revision as of 12:43, 16 March 2011
Contents
Install the package under R environmemnt
under Ubuntu install the R package:
sudo apt-get install r-base-core
type R to start the program:
R
type the command to download and install igraph as R library:
install.packages("igraph")
Starting with the package
don't forget to load the library:
library('igraph')
parse a network file in Pajek format and access to its memory representation through the variable 'g':
g <- read.graph(file='/your/path/.../file.net', format='pajek')
list vertex and their attributes of graph g:
V(g) # the list of all nodes
V(g)[0] # access the first node
list.vertex.attributes(g) # the list of all vertex attributes names
V(g)[0]$id # access the attribute 'id' for the first node
list edges/arcs and their attributes of graph g:
E(g) # the list of all edges/arcs E(g)[0] # access the first edge/arc list.edge.attributes(g) # the list of all edges/arcs attributes names E(g)[0]$weight # access the attribute 'weight' for the first node
draw the graph:
plot(g) tkplot(g)
Macroscopic Analysis
graph size (number of nodes): length(V(g))
number of edges: length(E(g))
diameter: diameter(g)
mean degree:
density:
giant component size (the largest connected component size):
assortativity:
Node Centrality Analysis
degree: degree(g)
closeness: closeness(g)
betweenness: betweenness(g)
eigenvector/pagerank: evcent(g)
Clustering Techniques
connected components: clusters(g)
edge betweenness:
walktrap:
spinglass:
References and further informations
Igraph/R documentation: http://igraph.sourceforge.net/doc/R/00Index.html
