Nothing
R/measures.R
In memnet: Network Tools for Memory Research
#' Retrieve largest component
#'
#' Retrieves the largest component. In case of equally sized
#' components the first is component is retrieved.
#'
#' @param adj numeric matrix representing the adjacency matrix.
#' @param weights numeric vector of edge weights. Optional.
#' @param mode character, either \code{"directed"} or \code{"undirected"},
#' specifying whether the network should be interepeted as directed
#' or undirected. Defaults to \code{"undirected"}.
#' @param igraph logical specifying whether the output should be of class
#' \code{"igraph"}.
#'
#' @return
#' A list containing the, now, named adjacency matrix and a numeric value
#' indicating the size of the largest component relative to
#' to the entire graph.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge list of fluency graph
#' edge_list = threshold_graph(animal_fluency[1:10])
#'
#' # get adjacency matrix
#' adj = edg_to_adj(edge_list)
#'
#' # get largest component
#' l_comp(adj)
#'
#' @export
l_comp = function(adj, weights = NULL, mode = 'undirected', igraph = FALSE){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# retrieve components
cmps = igraph::components(g)
# determine max component size
msiz = max(cmps$csize)
# Select nodes of max component
icmp = which.max(cmps$csize)
ncmp = (1:length(cmps$membership))[cmps$membership == icmp]
# component
comp = igraph::induced_subgraph(g,ncmp)
# return igraph
if(igraph == TRUE){
out = list("adj" = comp,
"rel_size" = msiz / igraph::vcount(g))
return(out)
}
# return largest component and size of largest component
out = list("adj" = igraph::get.adjacency(comp, sparse = FALSE),
"rel_size" = msiz / igraph::vcount(g))
out
}
#' Maximum difference between cumulative degree distribution
#'
#' Determines the maximum difference between the cumulative
#' degree distributions of two graphs
#'
#' @inheritParams l_comp
#' @param adj_1 numeric matrix representing the adjacency matrix of graph 1.
#' @param adj_2 numeric matrix representing the adjacency matrix of graph 2.
#' @param weights_1,weights_2 numeric vector of edge weights for network 1 and
#' 2, respectively. Optional.
#' @param mode_1,mode_2 character, either \code{"directed"} or \code{"undirected"},
#' specifying whether network 1 and 2 should be interepeted as directed
#' or undirected, respectively. Defaults to \code{"undirected"}.
#'
#' @return
#' Distance between cumulative degree distributions.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get max degree distance
#' k_dist(adj_1, adj_2)
#'
#' @export
k_dist = function(adj_1, adj_2,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
# get degree distance
da = igraph::degree.distribution(g1)
dg = igraph::degree.distribution(g2)
k_dist = trm(da, dg)
# out
names(k_dist) = 'k_dist'
k_dist
}
#' Average local clustering (alc) coefficient
#'
#' Computes the uncorrected or corrected average local clustering coffiecient.
#'
#' The uncorrected clustering coefficent is computed according to Watts &
#' Strogatz (1998). The corrected clustering coefficient normalizes the
#' uncorrected one by the \code{average degree / n nodes}, i.e., the expected
#' average local clustering for an Erdös-Renyi random graph.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param types character. Either \code{"uncorrected"} or \code{"corrected"}, or
#' a vector containing both.
#'
#' @return the corrected local clustering coefficient and/or
#' the uncorrected clustering coefficient.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get local average clustering coefficient
#' alc(adj)
#'
#' # get corrected local average clustering coefficient
#' alc(adj, types = 'corrected')
#'
#' @export
alc = function(adj, types = 'uncorrected', weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
res = c()
cc = igraph::transitivity(g, type = 'localaverage')
for(type in types){
if(type == 'uncorrected'){
res = c(res, cc)
names(res)[length(res)] = 'cc'
} else if(type == 'corrected'){
nv = igraph::vcount(g)
ne = igraph::ecount(g)
k = ne / nv
res = c(res, cc / (k / nv))
names(res)[length(res)] = 'cc_c'
}
}
res
}
#' Average shortest path length (aspl)
#'
#' Computes the average shortest path length (aspl) using \code{igraph}'s automatic
#' method.
#'
#' Per default the uncorrected apsl is computed. Otherwise, the uncorrected aspl
#' is normalized by \code{log(n nodes)/log(average degree)}, i.e., the expected
#' average shortest path length for an Erdös-Renyi random graph.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param types character. Either \code{"uncorrected"} or \code{"corrected"}, or
#' a vector containing both.
#'
#' @return Local clustering coefficient.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get average shortest path length
#' aspl(adj)
#'
#' # get corrected average shortest path length
#' aspl(adj, types = 'corrected')
#'
#' @export
aspl = function(adj, types = 'uncorrected', weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# get aspl
res = c()
aspl = igraph::mean_distance(g, directed = mode == 'directed')
for(type in types){
if(type == 'uncorrected'){
res = c(res, aspl)
names(res)[length(res)] = 'aspl'
} else if(type == 'corrected'){
n_v = igraph::vcount(g)
n_e = igraph::ecount(g)
k = n_e / n_v
res = c(res, aspl / (log(n_v) / log(k)))
names(res)[length(res)] = 'aspl_c'
}
}
res
}
#' Network statistics
#'
#' Function computes various network measures including the size of the number of
#' nodes (|V|), the number of edges (|E|), the average degree (k),
#' uncorrected and corrected average local clustering (C, Cc), uncorrected and
#' corrected average shortest path lengths (L, Lc), the small-world index (S)
#' from Humphries & Gurney (2008), assortivity (A), and the size of largest
#' component (p) relative to the entire network.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param giant logical specifying whether the should be computed for the
#' largest component.
#'
#' @return A named numeric vector containing the computed network statistics.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get structural overview
#' network_stats(adj)
#'
#' @export
network_stats = function(adj, giant = FALSE, weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# get largest component
comp = l_comp(g, igraph = TRUE)
comp_p = comp[[2]]
if(giant == TRUE){
g = comp[[1]]
}
# compute measures
n_v = igraph::vcount(g)
n_e = igraph::ecount(g)
k = n_e / n_v
cc = igraph::transitivity(g, type = 'localaverage')
cc_c = cc / (k / n_v)
aspl = igraph::mean_distance(g, directed = mode == 'directed')
aspl_c = aspl / (log(n_v) / log(n_e))
smallw = cc_c / aspl_c
assor = igraph::assortativity_degree(g, directed = mode == 'directed')
res = c(n_v, n_e, k, cc, cc_c, aspl, aspl_c, smallw, assor, comp_p)
names(res) = c('|V|', '|E|', 'k',
'C', 'Cc',
'L', 'Lc',
'S', 'A',
'p')
res
}
#' Get common subgraph
#'
#' Function identifies and returns the common subgraphs of two networks.
#'
#' @inheritParams l_comp
#' @inheritParams k_dist
#'
#' @return List containing the two subgraphs.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get common subgraph
#' common_subgraphs(adj_1, adj_2)
#'
#' @export
common_subgraphs = function(adj_1, adj_2,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph 1
if(class(adj_1) != 'igraph'){
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
} else {
g1 = adj_1
}
# to igraph 2
if(class(adj_2) != 'igraph'){
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
} else {
g2 = adj_2
}
# assign names
if(is.null(igraph::V(g1)$name)) igraph::V(g1)$name = 1:nrow(adj_1)
if(is.null(igraph::V(g2)$name)) igraph::V(g2)$name = 1:nrow(adj_2)
# get common nodes
common_nodes = intersect(igraph::V(g1)$name,
igraph::V(g2)$name)
# common subgraph
g1_com = igraph::induced_subgraph(g1, common_nodes)
g2_com = igraph::induced_subgraph(g2, common_nodes)
# out
out = list(igraph::get.adjacency(g1_com, sparse = FALSE),
igraph::get.adjacency(g2_com, sparse = FALSE))
out
}
#' Common subgraph statistics
#'
#' Function identifies the common subgraphs of two networks and returns the
#' networks statistics using \link{network_stats}.
#'
#' @inheritParams l_comp
#' @inheritParams k_dist
#' @inheritParams network_stats
#'
#' @return List containing the two subgraphs.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get structural overview of both networks
#' common_subgraph_stats(adj_1, adj_2)
#'
#' @export
common_subgraph_stats = function(adj_1, adj_2,
giant = FALSE,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph 1
if(class(adj_1) != 'igraph'){
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
} else {
g1 = adj_1
}
# to igraph 2
if(class(adj_2) != 'igraph'){
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
} else {
g2 = adj_2
}
# assign names
if(is.null(igraph::V(g1)$name)) igraph::V(g1)$name = 1:nrow(adj_1)
if(is.null(igraph::V(g2)$name)) igraph::V(g2)$name = 1:nrow(adj_2)
# get common nodes
common_nodes = intersect(igraph::V(g1)$name,
igraph::V(g2)$name)
# common subgraph
g1_com = igraph::induced_subgraph(g1, common_nodes)
g2_com = igraph::induced_subgraph(g2, common_nodes)
# compute stats
res_1 = network_stats(g1_com)
res_2 = network_stats(g2_com)
# out
out = rbind(res_1, res_2, res_1 - res_2)
rownames(out) = c('adj_1','adj_2','adj_1-adj_2')
out
}
Try the memnet package in your browser
Any scripts or data that you put into this service are public.
memnet documentation built on May 2, 2019, 9:35 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#' Retrieve largest component
#'
#' Retrieves the largest component. In case of equally sized
#' components the first is component is retrieved.
#'
#' @param adj numeric matrix representing the adjacency matrix.
#' @param weights numeric vector of edge weights. Optional.
#' @param mode character, either \code{"directed"} or \code{"undirected"},
#' specifying whether the network should be interepeted as directed
#' or undirected. Defaults to \code{"undirected"}.
#' @param igraph logical specifying whether the output should be of class
#' \code{"igraph"}.
#'
#' @return
#' A list containing the, now, named adjacency matrix and a numeric value
#' indicating the size of the largest component relative to
#' to the entire graph.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge list of fluency graph
#' edge_list = threshold_graph(animal_fluency[1:10])
#'
#' # get adjacency matrix
#' adj = edg_to_adj(edge_list)
#'
#' # get largest component
#' l_comp(adj)
#'
#' @export
l_comp = function(adj, weights = NULL, mode = 'undirected', igraph = FALSE){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# retrieve components
cmps = igraph::components(g)
# determine max component size
msiz = max(cmps$csize)
# Select nodes of max component
icmp = which.max(cmps$csize)
ncmp = (1:length(cmps$membership))[cmps$membership == icmp]
# component
comp = igraph::induced_subgraph(g,ncmp)
# return igraph
if(igraph == TRUE){
out = list("adj" = comp,
"rel_size" = msiz / igraph::vcount(g))
return(out)
}
# return largest component and size of largest component
out = list("adj" = igraph::get.adjacency(comp, sparse = FALSE),
"rel_size" = msiz / igraph::vcount(g))
out
}
#' Maximum difference between cumulative degree distribution
#'
#' Determines the maximum difference between the cumulative
#' degree distributions of two graphs
#'
#' @inheritParams l_comp
#' @param adj_1 numeric matrix representing the adjacency matrix of graph 1.
#' @param adj_2 numeric matrix representing the adjacency matrix of graph 2.
#' @param weights_1,weights_2 numeric vector of edge weights for network 1 and
#' 2, respectively. Optional.
#' @param mode_1,mode_2 character, either \code{"directed"} or \code{"undirected"},
#' specifying whether network 1 and 2 should be interepeted as directed
#' or undirected, respectively. Defaults to \code{"undirected"}.
#'
#' @return
#' Distance between cumulative degree distributions.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get max degree distance
#' k_dist(adj_1, adj_2)
#'
#' @export
k_dist = function(adj_1, adj_2,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
# get degree distance
da = igraph::degree.distribution(g1)
dg = igraph::degree.distribution(g2)
k_dist = trm(da, dg)
# out
names(k_dist) = 'k_dist'
k_dist
}
#' Average local clustering (alc) coefficient
#'
#' Computes the uncorrected or corrected average local clustering coffiecient.
#'
#' The uncorrected clustering coefficent is computed according to Watts &
#' Strogatz (1998). The corrected clustering coefficient normalizes the
#' uncorrected one by the \code{average degree / n nodes}, i.e., the expected
#' average local clustering for an Erdös-Renyi random graph.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param types character. Either \code{"uncorrected"} or \code{"corrected"}, or
#' a vector containing both.
#'
#' @return the corrected local clustering coefficient and/or
#' the uncorrected clustering coefficient.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get local average clustering coefficient
#' alc(adj)
#'
#' # get corrected local average clustering coefficient
#' alc(adj, types = 'corrected')
#'
#' @export
alc = function(adj, types = 'uncorrected', weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
res = c()
cc = igraph::transitivity(g, type = 'localaverage')
for(type in types){
if(type == 'uncorrected'){
res = c(res, cc)
names(res)[length(res)] = 'cc'
} else if(type == 'corrected'){
nv = igraph::vcount(g)
ne = igraph::ecount(g)
k = ne / nv
res = c(res, cc / (k / nv))
names(res)[length(res)] = 'cc_c'
}
}
res
}
#' Average shortest path length (aspl)
#'
#' Computes the average shortest path length (aspl) using \code{igraph}'s automatic
#' method.
#'
#' Per default the uncorrected apsl is computed. Otherwise, the uncorrected aspl
#' is normalized by \code{log(n nodes)/log(average degree)}, i.e., the expected
#' average shortest path length for an Erdös-Renyi random graph.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param types character. Either \code{"uncorrected"} or \code{"corrected"}, or
#' a vector containing both.
#'
#' @return Local clustering coefficient.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get average shortest path length
#' aspl(adj)
#'
#' # get corrected average shortest path length
#' aspl(adj, types = 'corrected')
#'
#' @export
aspl = function(adj, types = 'uncorrected', weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# get aspl
res = c()
aspl = igraph::mean_distance(g, directed = mode == 'directed')
for(type in types){
if(type == 'uncorrected'){
res = c(res, aspl)
names(res)[length(res)] = 'aspl'
} else if(type == 'corrected'){
n_v = igraph::vcount(g)
n_e = igraph::ecount(g)
k = n_e / n_v
res = c(res, aspl / (log(n_v) / log(k)))
names(res)[length(res)] = 'aspl_c'
}
}
res
}
#' Network statistics
#'
#' Function computes various network measures including the size of the number of
#' nodes (|V|), the number of edges (|E|), the average degree (k),
#' uncorrected and corrected average local clustering (C, Cc), uncorrected and
#' corrected average shortest path lengths (L, Lc), the small-world index (S)
#' from Humphries & Gurney (2008), assortivity (A), and the size of largest
#' component (p) relative to the entire network.
#'
#' @inheritParams l_comp
#' @param adj numeric matrix representing the adjacency matrix.
#' @param giant logical specifying whether the should be computed for the
#' largest component.
#'
#' @return A named numeric vector containing the computed network statistics.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list = threshold_graph(animal_fluency)
#'
#' # get adjacency matrices
#' adj = edg_to_adj(edge_list)
#'
#' # get structural overview
#' network_stats(adj)
#'
#' @export
network_stats = function(adj, giant = FALSE, weights = NULL, mode = 'undirected'){
# to igraph
if(class(adj) != 'igraph'){
g = igraph::graph_from_adjacency_matrix(adj, mode = mode)
if(!is.null(weights)) igraph::E(g)$weight = weights
} else {
g = adj
}
# get largest component
comp = l_comp(g, igraph = TRUE)
comp_p = comp[[2]]
if(giant == TRUE){
g = comp[[1]]
}
# compute measures
n_v = igraph::vcount(g)
n_e = igraph::ecount(g)
k = n_e / n_v
cc = igraph::transitivity(g, type = 'localaverage')
cc_c = cc / (k / n_v)
aspl = igraph::mean_distance(g, directed = mode == 'directed')
aspl_c = aspl / (log(n_v) / log(n_e))
smallw = cc_c / aspl_c
assor = igraph::assortativity_degree(g, directed = mode == 'directed')
res = c(n_v, n_e, k, cc, cc_c, aspl, aspl_c, smallw, assor, comp_p)
names(res) = c('|V|', '|E|', 'k',
'C', 'Cc',
'L', 'Lc',
'S', 'A',
'p')
res
}
#' Get common subgraph
#'
#' Function identifies and returns the common subgraphs of two networks.
#'
#' @inheritParams l_comp
#' @inheritParams k_dist
#'
#' @return List containing the two subgraphs.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get common subgraph
#' common_subgraphs(adj_1, adj_2)
#'
#' @export
common_subgraphs = function(adj_1, adj_2,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph 1
if(class(adj_1) != 'igraph'){
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
} else {
g1 = adj_1
}
# to igraph 2
if(class(adj_2) != 'igraph'){
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
} else {
g2 = adj_2
}
# assign names
if(is.null(igraph::V(g1)$name)) igraph::V(g1)$name = 1:nrow(adj_1)
if(is.null(igraph::V(g2)$name)) igraph::V(g2)$name = 1:nrow(adj_2)
# get common nodes
common_nodes = intersect(igraph::V(g1)$name,
igraph::V(g2)$name)
# common subgraph
g1_com = igraph::induced_subgraph(g1, common_nodes)
g2_com = igraph::induced_subgraph(g2, common_nodes)
# out
out = list(igraph::get.adjacency(g1_com, sparse = FALSE),
igraph::get.adjacency(g2_com, sparse = FALSE))
out
}
#' Common subgraph statistics
#'
#' Function identifies the common subgraphs of two networks and returns the
#' networks statistics using \link{network_stats}.
#'
#' @inheritParams l_comp
#' @inheritParams k_dist
#' @inheritParams network_stats
#'
#' @return List containing the two subgraphs.
#'
#' @examples
#'
#' # get fluency data
#' data(animal_fluency)
#'
#' # edge lists of fluency graphs
#' edge_list_1 = threshold_graph(animal_fluency[1:100])
#' edge_list_2 = threshold_graph(animal_fluency[101:200])
#'
#' # get adjacency matrices
#' adj_1 = edg_to_adj(edge_list_1)
#' adj_2 = edg_to_adj(edge_list_2)
#'
#' # get structural overview of both networks
#' common_subgraph_stats(adj_1, adj_2)
#'
#' @export
common_subgraph_stats = function(adj_1, adj_2,
giant = FALSE,
weights_1 = NULL, weights_2 = NULL,
mode_1 = 'undirected',
mode_2 = 'undirected'){
# to igraph 1
if(class(adj_1) != 'igraph'){
g1 = igraph::graph_from_adjacency_matrix(adj_1, mode = mode_1)
if(!is.null(weights_1)) igraph::E(g1)$weight = weights_1
} else {
g1 = adj_1
}
# to igraph 2
if(class(adj_2) != 'igraph'){
g2 = igraph::graph_from_adjacency_matrix(adj_2, mode = mode_2)
if(!is.null(weights_2)) igraph::E(g2)$weight = weights_2
} else {
g2 = adj_2
}
# assign names
if(is.null(igraph::V(g1)$name)) igraph::V(g1)$name = 1:nrow(adj_1)
if(is.null(igraph::V(g2)$name)) igraph::V(g2)$name = 1:nrow(adj_2)
# get common nodes
common_nodes = intersect(igraph::V(g1)$name,
igraph::V(g2)$name)
# common subgraph
g1_com = igraph::induced_subgraph(g1, common_nodes)
g2_com = igraph::induced_subgraph(g2, common_nodes)
# compute stats
res_1 = network_stats(g1_com)
res_2 = network_stats(g2_com)
# out
out = rbind(res_1, res_2, res_1 - res_2)
rownames(out) = c('adj_1','adj_2','adj_1-adj_2')
out
}
Try the memnet package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.