analyse_disparity_filter: Topological analysis of the disparity filter

In galanisl/DisparityFilter: Disparity filter for complex undirected weighted networks

Description

After disparity p-values have been obtained, applies different p-value thresholds to the given network and computes the fraction of remaining nodes, edges and size of the largest connected component from the perspective of the original network. It does the same for a filter based on weights (global filter). In addition, it recommends a disparity p-value to filter the network, keeping as many nodes as possible and the most significant edges.

Usage

analyse_disparity_filter(net, breaks = 100)

Arguments

net	igraph; The undirected weighted network to which the disparity filter was applied. It must contain a pval edge attribute.
breaks	integer; The length of the sequence of values between the maximum and minimum edge p-values.

Value

Data frame with the following columns:

threshold	The different p-value thresholds considered in the analysis.
N	The remaining fraction of nodes at each threshold.
L	The remaining fraction of links at each threshold.
W	The remaining fraction of the total weight at each threshold.
LCC_tot	Fraction of nodes from the original network in the largest connected component of the filtered network.
LCC_bb	Fraction of nodes from the filtered network in the largest connected component of the filtered network.
cc	Clustering coefficient of the filtered network at each threshold.
recommended	TRUE if the threshold is the recommended one to filter the network, FALSE otherwise.
filter	Whether the entry of the data frame corresponds to the Disparity or the Global filter.

Author(s)

Gregorio Alanis-Lobato [e-mail]

References

Serrano, M. A. et al. (2009) Extracting the multiscale backbone of complex weighted networks. PNAS 106(16).

Garcia-Perez, G. et al. (2016) The hidden hyperbolic geometry of international trade: World Trade Atlas 1870-2013. Scientific Reports 6(33441).

Examples

# Get disparity p-values for the edges of the included US Airports network
air_with_pval <- get_edge_disparity_pvals(net = air)
# Analyse the topology of the networks resulting from the application of 
# different disparity filters
analysis <- analyse_disparity_filter(net = air_with_pval, breaks = 100)

galanisl/DisparityFilter documentation built on May 16, 2019, 5:36 p.m.