inst/scripts/RSienaSNADescriptives.R

In RSienaTest: Siena - Simulation Investigation for Empirical Network Analysis

##################################################################################
###
### ---- RscriptSNADescriptives.R: a script for the introduction to RSiena -------
###
###                               version: May 8, 2014
##################################################################################
#
# Rscript01DataFormat.R is followed by
# RScriptSNADescriptives.R, code for descriptive analysis of the data, and
# Rscript02SienaVariableFormat.R, which formats data and specifies the model, and
# Rscript03SienaRunModel.R, which runs the model and estimates parameters
# Rscript04SienaBehaviour.R, which illustrates an example of analysing the
# coevolution of networks and behaviour
#
# The entire model fitting is summarised at the end of Rscript03SienaRunModel.R
# (without comments)
#
# This is an R script for getting started with RSiena, written by
# Robin Gauthier, Tom Snijders, Ruth Ripley, Johan Koskinen, and
# Paulina Preciado, with some examples borrowed from Christian Steglich.
# Lines starting with # are not processed by R but treated as comments.
# The script has a lot of explanation of R possibilities that will be
# familiar for readers well acquainted with R, and can be skipped by them.

# A visual inspection of the adjacency matrices can sometimes be useful.
# This will, for example, help in highlighting outliers with respect to
# outdegrees or indegrees, if there are any of such outliers.
# This requires package "sna":

        library( network )
        library( sna )

# For this script, you will need the data read and modified in the script
# Rscript01DataFormat.R. If you have already ran that script, you may
# load the required workspace:
load("WorkspaceRscript01.RData")

        net1 <- as.network( friend.data.w1 )
        net2 <- as.network( friend.data.w2 )
        net3 <- as.network( friend.data.w3 )
        plot.sociomatrix( net1,drawlab = F, diaglab = F, xlab = 'friendship t1' )
        plot.sociomatrix( net2,drawlab = F, diaglab = F, xlab = 'friendship t2' )
        plot.sociomatrix( net3,drawlab = F, diaglab = F, xlab = 'friendship t3' )

# The class,
	class( net1 )
# with attributes
	attributes( net1 )
# has special methods associated with it.
# while  plot( friend.data.w1 ) only produces a rather dull plot of
# the first two columns
#       plot( net1, xlab = 'friendship t1' )
# produces a nice sociogram
#
# Some further descriptives you can do for the data are plotting and
# calculating some statistics.

# add the attribute drink to the network object

        net1 %v% "drink" <- drink[ , 1 ]

# color the nodes by drink

        plot( net1, vertex.col = "drink", xlab = 'friendship t1' )


# Now let's color the nodes by drink and scale the vertex by degree of nodes!
#
# First calculate degree:

        deg <- rowSums( as.matrix( net1 ) )# NB:  rowSums() is defined for class matrix

# have a look at the degree distribution

        table( deg, useNA = 'always' )

# Now do the desired plot:

        plot( net1, vertex.col = "drink", vertex.cex = (deg + 1)/1.5 )

# ---- Plot the three waves of data --------------------------------------------

# Add drink to waves 2 and 3
        net2 %v% "drink" <- drink[ , 2 ]
        net3 %v% "drink" <- drink[ , 3 ]
      	deg2 <- rowSums( as.matrix( net2 ) )
      	deg3 <- rowSums( as.matrix( net3 ) )

# Create a set of panels ( 1 row by 3 columns, or 3 columns by 1 row)

        par( mfrow = c( 1, 3 ) )

# creating three plots after each other will place them in consecutive panels

        plot( net1, vertex.col = "drink", vertex.cex = (deg + 1)/1.5 )
        plot( net2, vertex.col = "drink", vertex.cex = (deg2 + 1)/1.5 )
        plot( net3, vertex.col = "drink", vertex.cex = (deg3 + 1)/1.5 )

# Each time we make a plot the coordinates move - because always
# the starting values are random. We can also save coordinates
# and use them for later plotting:

        par( mfrow = c( 1, 3 ) )
        coordin <-  plot( net1, vertex.col = "drink", vertex.cex = (deg +1 )/1.5 )
        plot( net2, coord = coordin, vertex.col = "drink", vertex.cex = (deg2 + 1)/1.5 )
        plot( net3, coord = coordin, vertex.col = "drink", vertex.cex = (deg3 + 1) /1.5 )

# To get coordinates based on all three waves: coordin <-  plot( net1 + net2 + net3 )
# For more plotting options, try the gplot function in the "sna" library

        ?gplot
        ?gplot.layout


# ---- Basic network statistics ------------------------------------------------

# The package "sna" can be used for a variety of descriptions and analyses.
# The following are examples.
# some important graph level statistics

        gden( net1 ) # density
        grecip( net1 ) # proportion of dyads that are symmetric
        grecip( net1, measure = "dyadic.nonnull" ) # reciprocity, ignoring the null dyads
        gtrans( net1 ) # transitivity

# dyad and triad census

        dyad.census( net1 )
        triad.census( net1 )

# out degree distribution (of course for a symmetric network outdegree=indegree)

        outdegree <- degree( net1, cmode = "outdegree" )
        outdegree #outgoing ties of each note

        hist( outdegree )
        quantile( outdegree )

# measures of connectivity and distance

        dist <- geodist(net1, inf.replace = Inf, count.paths = TRUE)
# calculate the geodesic distance (shortest path length) matrix
   	   dist$gd
# matrix of geodesic distances
	      dist$counts
# matrix containing the number of paths at each geodesic between each pair of vertices
        dist
        reach <- reachability( net1 )  # calculate the reachability matrix
        reach
################################################################################
###
### ---- PROCEED TO Rscript02SienaVariableFormat.R FOR PREPARING DATA FOR RSIENA -
###
################################################################################