added evcent and tests

.gitignore

@@ -35,3 +35,4 @@ vignettes/*.pdf
 # Shiny token, see https://shiny.rstudio.com/articles/shinyapps.html
 rsconnect/
 .Rproj.user
+inst/doc

DESCRIPTION

@@ -23,7 +23,10 @@ Suggests:
     testthat (>= 2.1.0),
     covr,
     ggplot2,
-    ggraph
+    ggraph,
+    knitr,
+    rmarkdown
 LinkingTo: 
     Rcpp,
     RcppArmadillo
+VignetteBuilder: knitr

NAMESPACE

@@ -4,13 +4,15 @@ export(as_adj_complex)
 export(as_adj_signed)
 export(as_complex_edges)
 export(as_incidence_complex)
+export(as_incidence_signed)
 export(as_signed_proj)
 export(as_unsigned_2mode)
 export(balance_score)
 export(complex_walks)
 export(count_complex_triangles)
 export(count_signed_triangles)
 export(degree_signed)
+export(eigen_centrality_signed)
 export(ggblock)
 export(ggsigned)
 export(graph_circular_signed)

R/centrality_indices.R

@@ -7,6 +7,21 @@
 #' @references Everett, M. and Borgatti, S. (2014) Networks containing negative ties. *Social Networks* 38 111-120
 #' @author David Schoch
 #' @importFrom Matrix t
+#' @examples
+#' library(igraph)
+#' A <- matrix(c(0,  1,  0,  1,  0,  0,  0, -1, -1,  0,
+#'               1,  0,  1, -1,  1, -1, -1,  0,  0,  0,
+#'               0,  1,  0,  1, -1,  0,  0,  0, -1,  0,
+#'               1, -1,  1,  0,  1, -1, -1,  0,  0,  0,
+#'               0,  1, -1,  1,  0,  1,  0, -1,  0, -1,
+#'               0, -1,  0, -1,  1,  0,  1,  0,  1, -1,
+#'               0, -1,  0, -1,  0,  1,  0,  1, -1,  1,
+#'              -1,  0,  0,  0, -1,  0,  1,  0,  1,  0,
+#'              -1,  0, -1,  0,  0,  1, -1,  1,  0,  1,
+#'               0,  0,  0,  0, -1, -1,  1,  0,  1,  0), 10, 10)
+
+#'g <- igraph::graph_from_adjacency_matrix(A,"undirected",weighted = "sign")
+#'pn_index(g)
 #' @export
 
 pn_index <- function(g,mode=c("all","in","out")){
@@ -95,6 +110,61 @@ degree_signed <- function(g,mode=c("all","in","out"), type = c("pos","neg","rati
                   ratio = Matrix::colSums(P)/(Matrix::colSums(P)+Matrix::colSums(N)),
                   net   = Matrix::colSums(P)-Matrix::colSums(N)
     )
-    res
+    return(res)
   }
 }
+
+#' @title Signed Eigenvector centrality
+#' @description Calcul
+#' @details Note that, with negative values, the adjacency matrix may not have a dominant eigenvalue.
+#' This means it is not clear which eigenvector should be used. In addition it is possible for the adjacency matrix to have repeated #'eigenvalues and hence multiple linearly independent eigenvectors. In this case certain centralities can be arbitrarily assigned. The function returns an error if this is the case.
+#' @param g igraph object. Must have a "sign" edge attribute.
+#' @param scale Logical scalar, whether to scale the result to have a maximum score of one. If no scaling is used then the result vector is the same as returned by `eigen()`.
+#' @return centrality scores as numeric vector.
+#' @references
+#' Bonacich, P. and Lloyd, P. (2004). "Calculating Status with Negative Relations." *Social Networks* 26 (4): 331–38.
+#'
+#' Everett, M. and Borgatti, S.P. (2014). "Networks Containing Negative Ties." *Social Networks* 38: 111–20.
+#'
+#' @author David Schoch
+#' @examples
+#' library(igraph)
+#' # example for network without dominant eigenvalue (from Everett&Borgatti)
+#'
+#' A <- matrix(c( 0,  1,  1, -1,  0,  0, -1,  0,  0,
+#'                1,  0,  1,  0, -1,  0,  0, -1,  0,
+#'                1,  1,  0,  0,  0, -1,  0,  0, -1,
+#'               -1,  0,  0,  0,  1,  1, -1,  0,  0,
+#'                0, -1,  0,  1,  0,  1,  0, -1,  0,
+#'                0,  0, -1,  1,  1,  0,  0,  0, -1,
+#'               -1,  0,  0, -1,  0,  0,  0,  1,  1,
+#'                0, -1,  0,  0, -1,  0,  1,  0,  1,
+#'                0,  0, -1,  0,  0, -1,  1,  1, 0), 9, 9)
+#' g <- graph_from_adjacency_matrix(A,"undirected",weighted = "sign")
+#'
+#' # eigen_centrality_signed(g)
+#'
+#' @export
+
+eigen_centrality_signed <- function(g, scale = TRUE){
+  if (!igraph::is_igraph(g)) {
+    stop("Not a graph object")
+  }
+  if(!"sign"%in%igraph::edge_attr_names(g)){
+    stop("network does not have a sign edge attribute")
+  }
+  A <- as_adj_signed(g,sparse = TRUE)
+  sA <- eigen(A)
+  evals <- round(sA$values,8)
+  max_evals <- which(abs(evals)==max(abs(evals)))
+
+  if(length(max_evals)!=1){
+    stop("no dominant eigenvalue exists")
+  } else{
+    evcent <- sA$vectors[,max_evals]
+  }
+
+  if(scale) evcent <- evcent/max(evcent)
+
+  return(evcent)
+}

R/complex_matrices.R

@@ -18,6 +18,28 @@ as_adj_signed <- function(g,sparse = FALSE){
   igraph::as_adj(g,type = "both",attr = "sign", sparse = sparse)
 }
 
+#' Convert a signed two-mode network to a signed matrix
+#'
+#' This function returns the incidence matrix for a signed two-mode network.
+#'
+#' @param g igraph object (bipartite). Must have a "sign" edge attribute.
+#' @param sparse 	Logical scalar, whether to return the result as a sparse matrix. The Matrix package is required for sparse matrices.
+#' @return signed incidence matrix
+#' @export
+
+as_incidence_signed <- function(g,sparse = FALSE){
+  if (!igraph::is_igraph(g)) {
+    stop("Not a graph object")
+  }
+  if(!"sign"%in%igraph::edge_attr_names(g)){
+    stop("network does not have a sign edge attribute")
+  }
+  if(!"type"%in%igraph::vertex_attr_names(g)){
+    stop("network must have a type vertex attribute")
+  }
+  igraph::as_incidence_matrix(g,attr = "sign", sparse = sparse)
+}
+
 
 #' Convert a signed graph to a complex adjacency matrix
 #'

tests/testthat/test-centrality_indices.R

@@ -63,3 +63,44 @@ test_that("pn index works",{
   igraph::E(g)$sign <- 1
   expect_equal(pn_index(g,mode="in"),c(2, 2, 2, 2, 2))
 })
+
+test_that("evcent not graph error works",{
+  expect_error(eigen_centrality_signed(g = 5))
+})
+
+test_that("evcent no sign error works",{
+  g <- igraph::graph.full(5)
+  expect_error(eigen_centrality_signed(g))
+})
+
+test_that("evcent not dominant works",{
+  A <- matrix(c( 0,  1,  1, -1,  0,  0, -1,  0,  0,
+                 1,  0,  1,  0, -1,  0,  0, -1,  0,
+                 1,  1,  0,  0,  0, -1,  0,  0, -1,
+                -1,  0,  0,  0,  1,  1, -1,  0,  0,
+                 0, -1,  0,  1,  0,  1,  0, -1,  0,
+                 0,  0, -1,  1,  1,  0,  0,  0, -1,
+                -1,  0,  0, -1,  0,  0,  0,  1,  1,
+                 0, -1,  0,  0, -1,  0,  1,  0,  1,
+                 0,  0, -1,  0,  0, -1,  1,  1, 0), 9, 9)
+  g <- igraph::graph_from_adjacency_matrix(A,"undirected",weighted = "sign")
+  expect_error(eigen_centrality_signed(g))
+})
+
+test_that("evcent works",{
+  g <- igraph::graph.full(5)
+  igraph::E(g)$sign <- 1
+  igraph::E(g)$sign[1] <- -1
+  ev <- round(eigen_centrality_signed(g, scale = TRUE),8)
+  ev_true <- c(0.68614066, 0.68614066, 1, 1, 1)
+  expect_equal(ev,ev_true)
+})
+
+test_that("evcent no scale works",{
+  g <- igraph::graph.full(5)
+  igraph::E(g)$sign <- 1
+  igraph::E(g)$sign[1] <- -1
+  ev <- round(eigen_centrality_signed(g, scale = FALSE),8)
+  ev_true <- c(0.34560347, 0.34560347, 0.50369186, 0.50369186, 0.50369186)
+  expect_equal(ev,ev_true)
+})

tests/testthat/test-complex_matrices.R

@@ -1,9 +1,33 @@
-test_that("complex: error not graph works", {
-  expect_error(as_adj_complex(g=5))
+test_that("incidence: error not graph works", {
+  expect_error(as_incidence_signed(g=5))
 })
 
-test_that("complex: error directed works", {
-  expect_error(as_adj_complex(igraph::graph.full(3,directed=TRUE)))
+test_that("incidence: error no sign works", {
+  expect_error(as_incidence_signed(igraph::graph.full(3,directed=FALSE)))
+})
+
+test_that("incidence: error no type works", {
+  g <- igraph::graph.full(3,directed=FALSE)
+  igraph::E(g)$sign <- 1
+  expect_error(as_incidence_signed(g))
+})
+
+test_that("signed incidence works", {
+  A_true <- matrix(c(1,1,1,-1,-1,-1,-1,
+                1,1,1,-1,-1,-1,-1,
+                1,1,1,-1,-1,-1,-1,
+                -1,-1,-1,1,1,1,1,
+                -1,-1,-1,1,1,1,1),5,7,byrow = T)
+  rownames(A_true) <- letters[1:5]
+  colnames(A_true) <- 1:7
+  g <- igraph::graph_from_incidence_matrix(A_true,weighted = "sign")
+  A <- as_incidence_signed(g)
+  expect_equal(A,A_true)
+})
+
+
+test_that("complex: error not graph works", {
+  expect_error(as_adj_complex(g=5))
 })
 
 test_that("complex: error directed works", {