R: Simulate a bipartite network using sequential importance...

simulate.sisnetwork {networksis}

R Documentation

Simulate a bipartite network using sequential importance sampling

Description

The method simulate.sisnetwork simulates graphs with the same marginals as the passed network or as the rows and columns specified in a sisnetwork object through sequential importance sampling. That is, the degrees of the nodes are fixed and specified.

Usage

## S3 method for class 'sisnetwork'
simulate(object, nsim = 1, seed = NULL, save.networks = FALSE, ...)

Arguments

`object`	Either a `network` object or `sisnetwork` object. If a `sisnetwork` object, this should be a list with components `row` and `col` to specify the row and column degrees. These are the degrees of the type 1 and type 2 nodes, respectively.
`nsim`	Number of networks to be randomly drawn from the set of all networks.
`seed`	Seed for random number generator.
`save.networks`	If this is `TRUE`, the sampled networks are returned. Otherwise only the last network is returned.
`...`	Further arguments passed to or used by methods.

Details

A sample of networks is randomly drawn from the space of networks with the same degrees for each node.

Value

simulate.sisnetwork returns an object of class network.series, that is a list consisting of the following elements:

`networks`	The vector of simulated networks.
`log.prob`	The vector of the logarithm of the probability of being sampled.
`log.graphspace.size`	The logarithm of the mean estimate of the number of graphs in the graph space.
`log.graphspace.SE`	The logarithm of the standard error of the mean estimate of the number of graphs in the graph space.
`log.graphspace.size.lne`	The logarithm of the lognormal-based estimate of the number of graphs in the graph space.
`log.graphspace.SE.lne`	The logarithm of the standard error of the lognormal-based estimate of the number of graphs in the graph space.

Examples

bipartite.graph <- matrix(c(1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0), nrow = 3, byrow = TRUE)
example.net <- network(bipartite.graph)

# Specify the set to which each node belongs
example.net %v% "set" <- c(rep(1, 3),rep(2, 4))

# Simulate 100 graphs with the same marginals as 'example.net'
sim <- simulate.sisnetwork(example.net, nsim = 100)

# Estimated graph space size and SE
exp(sim$log.graphspace.size)
exp(sim$log.graphspace.SE)

# Darwin's finches example
data(finch)

sim <- simulate.sisnetwork(finch, nsim = 100, save.networks = TRUE)

# Calculate importance weights from the graph probabilities
importance.weights <- 1 / exp(sim$log.prob)
hist(importance.weights, breaks = 25, xlab = "Inverse Graph Probability", main="")

# Calculate Sanderson's \bar{S}^2
s.bar.squared.vec <- rep(0, 100)

for(i in 1 : 100)
{
   # Extract simulated bipartite graphs
   new.graph <- as.matrix.network(sim$networks[[i]])

   # Calculate custom graph statistic
   s.bar.squared.vec[i] <- (sum((new.graph %*% t(new.graph)) ^ 2) -
   sum(diag((new.graph %*% t(new.graph)) ^ 2))) / (13 * 12)
}

[Package networksis version 2.1-3 Index]