Save centrality data as csv

Author: DanielThedie

NAMESPACE

@@ -3,9 +3,7 @@
 S3method(add,Report)
 S3method(add_figure,Report)
 S3method(export_pdf,Report)
-S3method(get_betweenness,Interactions)
 S3method(get_centrality,Interactions)
-S3method(get_closeness,Interactions)
 S3method(get_cohesion,Interactions)
 S3method(get_communities,Interactions)
 S3method(get_density,Interactions)
@@ -15,6 +13,7 @@ S3method(get_transitivity,Interactions)
 S3method(plot_graph,Interactions)
 S3method(plot_top_authors,Interactions)
 S3method(print,Report)
+S3method(save_centrality_data,Interactions)
 S3method(save_md,Report)
 export(Interactions)
 export(Report)

R/Assemble_report.r

@@ -74,6 +74,9 @@ assemble_report <- function(config) {
       fig_caption = "Direct connections between the 15 most central authors"
     )
 
+    # Save centrality data as csv
+    save_centrality_data(interactions, paste0(paths$output, "/centrality_data_", date_range, ".csv"))
+
     # Save report as markdown
     save_md(report, paste0(paths$output, "/Report_", date_range, ".md"))
 

R/Interactions.plots.R

@@ -3,7 +3,7 @@ plot_graph.Interactions <- function(interactions, output_file = "output/graph.pn
   comm <- get_communities(interactions)
   colors <- grDevices::rainbow(length(unique(comm$membership)), alpha = 0.4)
 
-  centrality <- get_centrality(interactions)
+  centrality <- get_centrality(interactions)$degree
   vertex_size <- 1 + (centrality / max(centrality)) * 15
 
   grDevices::png(filename = output_file, width = 2500, height = 1800, res = 360)
@@ -41,7 +41,7 @@ plot_graph <- function(interactions, output_file) {
 #' @export
 plot_top_authors.Interactions <- function(interactions, n = 10, output_file = "output/top_authors.png") {
   # Identify the top n authors based on centrality
-  centrality <- get_centrality(interactions)
+  centrality <- get_centrality(interactions)$degree
   top_authors <- order(centrality, decreasing = TRUE)[1:n]
 
   # Create a subgraph with the top n authors

R/Interactions_class.r

@@ -93,31 +93,16 @@ get_transitivity <- function(interactions) {
 
 #' @export
 get_centrality.Interactions <- function(interactions) {
-  return(igraph::degree(interactions$graph))
+  degree <- igraph::degree(interactions$graph)
+  closeness <- igraph::closeness(interactions$graph)
+  betweenness <- igraph::betweenness(interactions$graph)
+  return(list(degree = degree, closeness = closeness, betweenness = betweenness))
 }
 
 get_centrality <- function(interactions) {
   UseMethod("get_centrality", interactions)
 }
 
-#' @export
-get_betweenness.Interactions <- function(interactions) {
-  return(igraph::betweenness(interactions$graph))
-}
-
-get_betweenness <- function(interactions) {
-  UseMethod("get_betweenness", interactions)
-}
-
-#' @export
-get_closeness.Interactions <- function(interactions) {
-  return(igraph::closeness(interactions$graph))
-}
-
-get_closeness <- function(interactions) {
-  UseMethod("get_closeness", interactions)
-}
-
 #' @export
 get_diameter.Interactions <- function(interactions) {
   return(igraph::diameter(interactions$graph, directed = interactions$directed, weights = NULL))
@@ -139,7 +124,7 @@ get_communities <- function(interactions) {
 #' @export
 get_most_central_per_community.Interactions <- function(interactions) {
   comm <- get_communities(interactions)
-  centrality <- get_centrality(interactions)
+  centrality <- get_centrality(interactions)$degree
   most_central_authors <- sapply(unique(comm$membership), function(group) {
     group_vertices <- which(comm$membership == group)
     group_centrality <- centrality[group_vertices]
@@ -152,3 +137,19 @@ get_most_central_per_community.Interactions <- function(interactions) {
 get_most_central_per_community <- function(interactions) {
   UseMethod("get_most_central_per_community", interactions)
 }
+
+#' @export
+save_centrality_data.Interactions <- function(interactions, output_file = "output/centrality_data.csv") {
+  centrality <- get_centrality(interactions)
+  output_data <- data.frame(
+    degree = centrality$degree,
+    closeness = centrality$closeness,
+    betweenness = centrality$betweenness
+  )
+  output_data <- output_data[statnet.common::order(-output_data$degree), ]
+  utils::write.csv(output_data, output_file, row.names = TRUE)
+}
+
+save_centrality_data <- function(interactions, output_file) {
+  UseMethod("save_centrality_data", interactions)
+}

R/Report_text.r

@@ -101,33 +101,31 @@ density_transitivity <- function(interactions) {
 #' @export
 centrality_metrics <- function(interactions) {
   centrality <- get_centrality(interactions)
-  betweenness <- get_betweenness(interactions)
-  closeness <- get_closeness(interactions)
   diameter <- get_diameter(interactions)
   text <- paste0("\n# Node Centrality and Centralization",
                  "\n## Degree Centrality",
                  "\nDegree centrality measures the number of direct connections each node has in the network.
                  Nodes with higher degree centrality are likely to be influential or well-connected individuals/entities.",
-                 "\n\n- **The least connected node has (Min)**: ", round(min(centrality)),
-                 "\n- **The most connected node has (Max)**: ", round(max(centrality)),
-                 "\n- **On average, nodes have about (Mean)**: ", round(mean(centrality)),
-                 "\n- **Half of the nodes have more connection than (Median)**: ", round(median(centrality)),
+                 "\n\n- **The least connected node has (Min)**: ", round(min(centrality$degree)),
+                 "\n- **The most connected node has (Max)**: ", round(max(centrality$degree)),
+                 "\n- **On average, nodes have about (Mean)**: ", round(mean(centrality$degree)),
+                 "\n- **Half of the nodes have more connection than (Median)**: ", round(median(centrality$degree)),
 
                  "\n\n## Betweenness Centrality",
                  "\nBetweenness centrality quantifies how often a node acts as a bridge in the shortest paths between other nodes.
                  Nodes with high betweenness are critical connectors in the network. Removing them could significantly disrupt the flow of information.",
-                 "\n\n- **Min**: ", round(min(betweenness)),
-                 "\n- **The node with the highest value is a critical bridge in the network, controlling the flow of information. with Max**: ", round(max(betweenness)),
-                 "\n- **Mean**: ", round(mean(betweenness)),
-                 "\n- **Median**: ", round(median(betweenness)),
+                 "\n\n- **Min**: ", round(min(centrality$betweenness)),
+                 "\n- **The node with the highest value is a critical bridge in the network, controlling the flow of information. with Max**: ", round(max(centrality$betweenness)),
+                 "\n- **Mean**: ", round(mean(centrality$betweenness)),
+                 "\n- **Median**: ", round(median(centrality$betweenness)),
 
                  "\n\n## Closeness Centrality",
                  "\nCloseness centrality measures how quickly a node can access other nodes in the network.
                  Nodes with high closeness centrality are in advantageous positions to disseminate information quickly.",
-                 "\n\n- **The least central node has a (Min) avarage path to reach other nodes**: ", round(min(closeness), digits = 5),
-                 "\n- **Max**: ", round(max(closeness), digits = 5),
-                 "\n- **Mean**: ", round(mean(closeness), digits = 5),
-                 "\n- **Median**: ", round(median(closeness), digits = 5),
+                 "\n\n- **The least central node has a (Min) avarage path to reach other nodes**: ", round(min(centrality$closeness), digits = 5),
+                 "\n- **Max**: ", round(max(centrality$closeness), digits = 5),
+                 "\n- **Mean**: ", round(mean(centrality$closeness), digits = 5),
+                 "\n- **Median**: ", round(median(centrality$closeness), digits = 5),
                  "\n\n## Network diameter",
                  "\nThe network diameter is the maximum distance (in terms of edges or steps) required to connect any two nodes in the network through the shortest possible path. ",
                  "\n\n- In our network: **the diameter is ", diameter, "**, it indicates that the farthest two nodes in your co-authorship network can be linked by the shortest path of ", diameter, " steps. ",

tests/testthat/test-Assemble-report.R

@@ -83,14 +83,29 @@ test_that("assemble_report works correctly", {
 
   mock_save_md <- function(report, file_path) {
     md_path <- paste0(folder, "/Report_2000-2001.md")
-    writeLines("Mock PDF Content", con = md_path)
+    writeLines("Mock MD Content", con = md_path)
   }
 
   mock_export_pdf <- function(report, input_file, output_file) {
     pdf_path <- paste0(folder, "/Report_2000-2001.pdf")
     writeLines("Mock PDF Content", con = pdf_path)
   }
 
+  mock_save_centrality_data <- function(interactions, output_path) {
+    csv_path <- paste0(folder, "/centrality_data_2000-2001.csv")
+    writeLines("Mock CSV Content", con = csv_path)
+  }
+
+  mock_get_author_stats <- function(data, author_column_name = "Author", delimiter = ";") {
+    list(
+      sum = 3,
+      average = 2,
+      median = 2,
+      min = 1,
+      max = 3
+    )
+  }
+
   # Use with_mock to override the functions
   testthat::with_mocked_bindings(
     import_csv_data = mock_import_csv_data,
@@ -109,6 +124,8 @@ test_that("assemble_report works correctly", {
     plot_top_authors = mock_plot_top_authors,
     save_md = mock_save_md,
     export_pdf = mock_export_pdf,
+    save_centrality_data = mock_save_centrality_data,
+    get_author_stats = mock_get_author_stats,
     {
       # Run the assemble_report function with the mock config
       assemble_report(config)

tests/testthat/test-Interactions.R

@@ -36,15 +36,10 @@ test_that("transitivity works", {
 })
 
 test_that("centrality works", {
-  expect_equal(mean(get_centrality(interactions)), 3)
-})
-
-test_that("betweenness works", {
-  expect_equal(round(mean(get_betweenness(interactions)), digits = 3), 0.167)
-})
-
-test_that("closeness works", {
-  expect_equal(round(mean(get_closeness(interactions)), digits = 3), 0.292)
+  centrality <- get_centrality(interactions)
+  expect_equal(mean(centrality$degree), 3)
+  expect_equal(round(mean(centrality$betweenness), digits = 3), 0.167)
+  expect_equal(round(mean(centrality$closeness), digits = 3), 0.292)
 })
 
 test_that("diameter works", {
@@ -60,3 +55,32 @@ test_that("most_central_authors works", {
   author_list <- get_most_central_per_community(interactions)
   expect_equal(author_list, "Author1")
 })
+
+test_that("save_centrality_data saves the correct centrality data to a CSV file", {
+  # Create a sample data frame
+  data <- data.frame(
+    Author = c("Author1;Author2", "Author2;Author3", "Author1;Author3"),
+    Year = c(2020, 2021, 2022)
+  )
+
+  # Create an Interactions object
+  interactions <- Interactions(data)
+
+  # Define the output file path
+  output_file <- tempfile(fileext = ".csv")
+
+  # Call the save_centrality_data function
+  save_centrality_data(interactions, output_file)
+
+  # Read the output file
+  output_data <- read.csv(output_file)
+
+  # Check that the output data has the correct columns
+  expect_true(all(c("degree", "closeness", "betweenness") %in% colnames(output_data)))
+
+  # Check that the output data is sorted by degree in descending order
+  expect_true(all(diff(output_data$degree) <= 0))
+
+  # Clean up the temporary file
+  unlink(output_file)
+})