Hands-on Exercise 7a

Modelling, Visualising and Analysing Network Data with R

Author

Yee Weicong Mark

Published

June 5, 2026

Modified

June 5, 2026

27 Chapter 27: Modelling, Visualising and Analysing Network Data with R

27.1 Overview

Network data describes entities and the connections between them. This chapter walks through the full workflow. A graph object is built from separate node and edge tables with tidygraph, drawn with ggraph, augmented with centrality and community metrics, then rendered interactively with visNetwork.

The dataset is the GAStech email corpus, containing two weeks of internal email between 55 employees of an oil exploration company. Each email is one edge from sender to recipient, each employee is one node with a department and title.

27.2 Getting Started

27.2.1 Installing and launching R packages

Nine packages drive this chapter.

igraph for the underlying graph data structure
tidygraph for tidy-style manipulation of graph objects
ggraph for ggplot2-style static network plots
visNetwork for interactive networks based on vis.js
lubridate and clock for time wrangling
graphlayouts for additional layout algorithms
concaveman and ggforce for drawing community hulls
tidyverse for data import and manipulation

pacman::p_load(igraph, tidygraph, ggraph,
               visNetwork, lubridate, clock,
               tidyverse, graphlayouts,
               concaveman, ggforce)

27.3 The Data

The data sets come from an oil exploration and extraction company. There are two data sets, one for the nodes and one for the edges.

27.3.1 The edges data

GAStech_email_edge-v2.csv consists of two weeks of 9,063 email correspondences between 55 employees. Each row carries source, target, SentDate, SentTime, Subject and MainSubject.

27.3.2 The nodes data

GAStech_email_node.csv consists of the 55 employees with id, label, Department and Title.

27.3.3 Importing network data from files

GAStech_nodes <- read_csv("data/GAStech_email_node.csv")
GAStech_edges <- read_csv("data/GAStech_email_edge-v2.csv")

27.3.4 Reviewing the imported data

glimpse(GAStech_edges)

Rows: 9,063
Columns: 8
$ source      <dbl> 43, 43, 44, 44, 44, 44, 44, 44, 44, 44, 44, 44, 26, 26, 26…
$ target      <dbl> 41, 40, 51, 52, 53, 45, 44, 46, 48, 49, 47, 54, 27, 28, 29…
$ SentDate    <chr> "6/1/2014", "6/1/2014", "6/1/2014", "6/1/2014", "6/1/2014"…
$ SentTime    <time> 08:39:00, 08:39:00, 08:58:00, 08:58:00, 08:58:00, 08:58:0…
$ Subject     <chr> "GT-SeismicProcessorPro Bug Report", "GT-SeismicProcessorP…
$ MainSubject <chr> "Work related", "Work related", "Work related", "Work rela…
$ sourceLabel <chr> "Sven.Flecha", "Sven.Flecha", "Kanon.Herrero", "Kanon.Herr…
$ targetLabel <chr> "Isak.Baza", "Lucas.Alcazar", "Felix.Resumir", "Hideki.Coc…

Things to learn from the code chunk above

glimpse() from dplyr shows column types and the first few values per column. The output reveals that SentDate was read as character rather than date, which would block any date-based filtering or sorting until corrected.

27.3.5 Wrangling time

GAStech_edges <- GAStech_edges %>%
  mutate(SendDate = dmy(SentDate)) %>%
  mutate(Weekday = wday(SentDate,
                        label = TRUE,
                        abbr = FALSE))

Things to learn from the code chunk above

dmy() parses a day-month-year character string into a true Date. wday() returns the weekday and label = TRUE returns it as an ordered factor with names rather than numbers. abbr = FALSE keeps full names (“Monday” rather than “Mon”). The ordering is meaningful because the factor levels follow calendar order rather than alphabetical.

27.3.6 Reviewing the revised date fields

The reformatted GAStech_edges data frame now has two additional columns. SendDate is a proper Date and Weekday is an ordered factor.

glimpse(GAStech_edges)

Rows: 9,063
Columns: 10
$ source      <dbl> 43, 43, 44, 44, 44, 44, 44, 44, 44, 44, 44, 44, 26, 26, 26…
$ target      <dbl> 41, 40, 51, 52, 53, 45, 44, 46, 48, 49, 47, 54, 27, 28, 29…
$ SentDate    <chr> "6/1/2014", "6/1/2014", "6/1/2014", "6/1/2014", "6/1/2014"…
$ SentTime    <time> 08:39:00, 08:39:00, 08:58:00, 08:58:00, 08:58:00, 08:58:0…
$ Subject     <chr> "GT-SeismicProcessorPro Bug Report", "GT-SeismicProcessorP…
$ MainSubject <chr> "Work related", "Work related", "Work related", "Work rela…
$ sourceLabel <chr> "Sven.Flecha", "Sven.Flecha", "Kanon.Herrero", "Kanon.Herr…
$ targetLabel <chr> "Isak.Baza", "Lucas.Alcazar", "Felix.Resumir", "Hideki.Coc…
$ SendDate    <date> 2014-01-06, 2014-01-06, 2014-01-06, 2014-01-06, 2014-01-0…
$ Weekday     <ord> Friday, Friday, Friday, Friday, Friday, Friday, Friday, Fr…

27.3.7 Wrangling attributes

Individual email records are too granular to visualise directly. Aggregating by sender, receiver and weekday produces an edge weight per pair per day.

GAStech_edges_aggregated <- GAStech_edges %>%
  filter(MainSubject == "Work related") %>%
  group_by(source, target, Weekday) %>%
    summarise(Weight = n()) %>%
  filter(source != target) %>%
  filter(Weight > 1) %>%
  ungroup()

Things to learn from the code chunk above

filter(MainSubject == "Work related") removes personal emails. group_by() and summarise(Weight = n()) count the emails per sender-receiver-weekday combination. filter(source != target) drops self-loops, and filter(Weight > 1) removes singleton pairs that would clutter the plot without revealing structure.

27.3.8 Reviewing the revised edges file

The aggregated edges file is the input to the graph object. Each row carries source, target, Weekday and Weight.

glimpse(GAStech_edges_aggregated)

Rows: 1,372
Columns: 4
$ source  <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
$ target  <dbl> 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 5, 5, 6,…
$ Weekday <ord> Sunday, Monday, Tuesday, Wednesday, Friday, Sunday, Monday, Tu…
$ Weight  <int> 5, 2, 3, 4, 6, 5, 2, 3, 4, 6, 5, 2, 3, 4, 6, 5, 2, 3, 4, 6, 5,…

27.4 Creating network objects using tidygraph

tidygraph treats a network as two tidy tibbles, one for nodes and one for edges, joined into a single tbl_graph object.

27.4.1 The tbl_graph object

Two functions of tidygraph create network objects. tbl_graph() builds one from scratch, and as_tbl_graph() converts existing igraph, network, dendrogram, hclust and data.tree objects.

27.4.2 The dplyr verbs in tidygraph

activate() switches between the node tibble and the edge tibble. All dplyr verbs applied to a tbl_graph operate on the currently active tibble. Inside one context, .N() reaches across to the node tibble, .E() to the edge tibble and .G() to the whole graph.

27.4.3 Using tbl_graph() to build tidygraph data model

GAStech_graph <- tbl_graph(nodes = GAStech_nodes,
                           edges = GAStech_edges_aggregated,
                           directed = TRUE)

Things to learn from the code chunk above

directed = TRUE preserves the asymmetry of email (sender to recipient is not the same as recipient to sender).

27.4.4 Reviewing the tidygraph graph object

GAStech_graph

# A tbl_graph: 54 nodes and 1372 edges
#
# A directed multigraph with 1 component
#
# Node Data: 54 × 4 (active)
      id label               Department     Title                               
   <dbl> <chr>               <chr>          <chr>                               
 1     1 Mat.Bramar          Administration Assistant to CEO                    
 2     2 Anda.Ribera         Administration Assistant to CFO                    
 3     3 Rachel.Pantanal     Administration Assistant to CIO                    
 4     4 Linda.Lagos         Administration Assistant to COO                    
 5     5 Ruscella.Mies.Haber Administration Assistant to Engineering Group Mana…
 6     6 Carla.Forluniau     Administration Assistant to IT Group Manager       
 7     7 Cornelia.Lais       Administration Assistant to Security Group Manager 
 8    44 Kanon.Herrero       Security       Badging Office                      
 9    45 Varja.Lagos         Security       Badging Office                      
10    46 Stenig.Fusil        Security       Building Control                    
# ℹ 44 more rows
#
# Edge Data: 1,372 × 4
   from    to Weekday Weight
  <int> <int> <ord>    <int>
1     1     2 Sunday       5
2     1     2 Monday       2
3     1     2 Tuesday      3
# ℹ 1,369 more rows

27.4.5 Interpreting the tidygraph graph object

The output is a tbl_graph with 54 nodes and 1,372 edges. The first six rows of the node tibble and the first three of the edge tibble are shown. The node tibble is active by default, which means subsequent dplyr verbs operate on it.

27.4.6 Changing the active object

The active tibble can be switched with activate(). Sorting the edge tibble by descending weight is a two-step pipeline.

GAStech_graph %>%
  activate(edges) %>%
  arrange(desc(Weight))

# A tbl_graph: 54 nodes and 1372 edges
#
# A directed multigraph with 1 component
#
# Edge Data: 1,372 × 4 (active)
    from    to Weekday   Weight
   <int> <int> <ord>      <int>
 1    40    41 Saturday      13
 2    41    43 Monday        11
 3    35    31 Tuesday       10
 4    40    41 Monday        10
 5    40    43 Monday        10
 6    36    32 Sunday         9
 7    40    43 Saturday       9
 8    41    40 Monday         9
 9    19    15 Wednesday      8
10    35    38 Tuesday        8
# ℹ 1,362 more rows
#
# Node Data: 54 × 4
     id label           Department     Title           
  <dbl> <chr>           <chr>          <chr>           
1     1 Mat.Bramar      Administration Assistant to CEO
2     2 Anda.Ribera     Administration Assistant to CFO
3     3 Rachel.Pantanal Administration Assistant to CIO
# ℹ 51 more rows

Things to learn from the code chunk above

activate(edges) switches the active tibble to edges so the subsequent arrange(desc(Weight)) reorders the edges from heaviest to lightest.

27.5 Plotting Static Network Graphs with ggraph package

ggraph extends ggplot2 for networks. Every graph is made up of nodes, edges and a layout that places nodes in 2-D space.

27.5.1 Plotting a basic network graph

ggraph(GAStech_graph) +
  geom_edge_link() +
  geom_node_point()

Things to learn from the code chunk above

ggraph() takes a tbl_graph or igraph object and sets the layout. geom_edge_link() draws straight-line edges between connected nodes, and geom_node_point() draws nodes as points. With no layout argument, the default is nicely, which picks a reasonable algorithm based on the graph type.

27.5.2 Changing the default network graph theme

g <- ggraph(GAStech_graph) +
  geom_edge_link(aes()) +
  geom_node_point(aes())

g + theme_graph()

Things to learn from the code chunk above

theme_graph() strips axes, grid lines and panel borders that carry no meaning in a network plot. It also switches the default font to Arial Narrow. set_graph_style() applies the same theme globally for a sequence of plots.

27.5.3 Changing the colouring of the plot

g <- ggraph(GAStech_graph) +
  geom_edge_link(aes(colour = 'grey50')) +
  geom_node_point(aes(colour = 'grey40'))

g + theme_graph(background = 'grey10',
                text_colour = 'white')

Things to learn from the code chunk above

background and text_colour in theme_graph() set the panel background and label colour together. A dark background can lift muted node colours, but it also reduces print friendliness.

27.5.4 Working with ggraph’s layouts

The layouts supported by ggraph are star, circle, nicely, dh, gem, graphopt, grid, mds, sphere, randomly, fr, kk, drl and lgl. Each chooses a different algorithm for placing nodes in 2-D space.

The grids below render the same GAStech_graph under every layout for visual comparison. Node colour encodes department and node size encodes degree.

V(GAStech_graph)$color <- as.numeric(as.factor(V(GAStech_graph)$Department))
V(GAStech_graph)$size  <- degree(GAStech_graph) / 3 + 3

par(mfrow = c(2, 3), mar = c(1, 1, 3, 1))
plot(GAStech_graph, layout = layout_as_star,    main = "layout_as_star",    vertex.label = NA)
plot(GAStech_graph, layout = layout_components, main = "layout_components", vertex.label = NA)
plot(GAStech_graph, layout = layout_in_circle,  main = "layout_in_circle",  vertex.label = NA)
plot(GAStech_graph, layout = layout_nicely,     main = "layout_nicely",     vertex.label = NA)
plot(GAStech_graph, layout = layout_on_grid,    main = "layout_on_grid",    vertex.label = NA)
plot(GAStech_graph, layout = layout_on_sphere,  main = "layout_on_sphere",  vertex.label = NA)

par(mfrow = c(3, 3), mar = c(1, 1, 3, 1))
plot(GAStech_graph, layout = layout_randomly,      main = "layout_randomly",      vertex.label = NA)
plot(GAStech_graph, layout = layout_with_dh,       main = "layout_with_dh",       vertex.label = NA)
plot(GAStech_graph, layout = layout_with_drl,      main = "layout_with_drl",      vertex.label = NA)
plot(GAStech_graph, layout = layout_with_fr,       main = "layout_with_fr",       vertex.label = NA)
plot(GAStech_graph, layout = layout_with_gem,      main = "layout_with_gem",      vertex.label = NA)
plot(GAStech_graph, layout = layout_with_graphopt, main = "layout_with_graphopt", vertex.label = NA)
plot(GAStech_graph, layout = layout_with_kk,       main = "layout_with_kk",       vertex.label = NA)
plot(GAStech_graph, layout = layout_with_lgl,      main = "layout_with_lgl",      vertex.label = NA)
plot(GAStech_graph, layout = layout_with_mds,      main = "layout_with_mds",      vertex.label = NA)

Things to learn from the code chunk above

V(GAStech_graph)$color and V(GAStech_graph)$size set node attributes that base R plot.igraph() reads. par(mfrow = c(2, 3)) arranges the panels in a 2 by 3 grid, and the same trick with c(3, 3) produces the 3 by 3 grid. The main argument labels each panel with the layout function name.

Reading the plot

Force-directed layouts like fr, kk, gem and graphopt produce visually similar results that emphasise cluster structure. Geometric layouts like star, circle, grid and sphere impose a fixed shape that ignores network structure. randomly is a useful baseline that shows how much the meaningful layouts contribute.

27.5.5 Fruchterman and Reingold layout

The Fruchterman-Reingold layout treats edges as springs and nodes as repelling particles, settling into a configuration where tightly connected groups cluster together.

g <- ggraph(GAStech_graph,
            layout = "fr") +
  geom_edge_link(aes()) +
  geom_node_point(aes())

g + theme_graph()

Things to learn from the code chunk above

layout = "fr" requests Fruchterman-Reingold. The result is non-deterministic, so two runs produce slightly different positions. Setting a random seed before plotting fixes the layout for reproducibility.

27.5.6 Modifying network nodes

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes()) +
  geom_node_point(aes(colour = Department,
                      size = 3))

g + theme_graph()

Things to learn from the code chunk above

geom_node_point() behaves like geom_point() in ggplot2. The colour = Department aesthetic encodes the department of each employee, and the constant size = 3 enlarges every node for readability.

Reading the plot

Departments cluster together because employees email their own department more than other departments. The visual grouping is an emergent property of the layout algorithm rather than something hard-coded by the colour mapping.

27.5.7 Modifying edges

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department),
                  size = 3)

g + theme_graph()

Things to learn from the code chunk above

width = Weight maps the edge thickness to the number of emails in that pair-weekday. scale_edge_width(range = c(0.1, 5)) clamps the actual rendered widths so the thinnest are still visible and the thickest do not dominate. alpha = 0.2 partially fixes the overplotting that arises when many edges cross.

27.6.1 Working with facet_edges()

set_graph_style()

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department),
                  size = 2)

g + facet_edges(~Weekday)

Things to learn from the code chunk above

facet_edges(~Weekday) keeps the node positions fixed across all panels and shows only the edges that occurred on each weekday. The unchanged node positions make the panels directly comparable across days.

Reading the plot

Weekday volume varies substantially. The differences between weekdays would be invisible in a single graph because all edges would superimpose into one tangle.

27.6.2 Adjusting the legend position

The legend is moved below the panels with theme(legend.position = 'bottom').

set_graph_style()

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department),
                  size = 2) +
  theme(legend.position = 'bottom')

g + facet_edges(~Weekday)

Things to learn from the code chunk above

legend.position = 'bottom' moves the legend below the panels to recover horizontal space.

27.6.3 A framed facet graph

set_graph_style()

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department),
                  size = 2)

g + facet_edges(~Weekday) +
  th_foreground(foreground = "grey80",  
                border = TRUE) +
  theme(legend.position = 'bottom')

Things to learn from the code chunk above

th_foreground() adds a grey panel background with border = TRUE drawing a thin frame around each facet. The borders help separate the panels visually when the layout pushes edges close to the panel edges.

27.6.4 Working with facet_nodes()

set_graph_style()

g <- ggraph(GAStech_graph,
            layout = "nicely") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department),
                  size = 2)

g + facet_nodes(~Department) +
  th_foreground(foreground = "grey80",  
                border = TRUE) +
  theme(legend.position = 'bottom')

Things to learn from the code chunk above

facet_nodes(~Department) shows the subgraph induced by each department, including only edges where both endpoints are in that department. Cross-department edges are dropped from every panel.

27.7 Network Metrics Analysis

27.7.1 Computing centrality indices

Centrality measures rank nodes by their structural importance in the network. The four canonical measures are degree (count of neighbours), betweenness (how often a node sits on shortest paths), closeness (inverse average distance to all other nodes) and eigenvector (importance weighted by neighbour importance).

g <- GAStech_graph %>%
  mutate(betweenness_centrality = centrality_betweenness()) %>%
  ggraph(layout = "fr") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department,
            size = betweenness_centrality))
g + theme_graph()

Things to learn from the code chunk above

mutate(betweenness_centrality = centrality_betweenness()) computes betweenness for each node and stores it as a new column on the active node tibble. Mapping it to size makes the most “bridge-like” employees appear as the largest points.

Reading the plot

The largest nodes are not the busiest emailers but the ones who connect otherwise distant parts of the network. They are typically managers or assistants whose role is to broker communication between departments.

27.7.2 Visualising network metrics

From ggraph 2.0 onwards, centrality measures can be computed inline inside an aes() call without a preceding mutate().

g <- GAStech_graph %>%
  ggraph(layout = "fr") +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(colour = Department,
                      size = centrality_betweenness()))
g + theme_graph()

Things to learn from the code chunk above

Calling centrality_betweenness() directly inside aes() removes the need to materialise the score on the graph object. The trade-off is that the score is recomputed on every plot, which matters for large graphs but not for this 54-node example.

27.7.3 Visualising Community

Community detection partitions nodes into clusters where within-cluster ties are denser than between-cluster ties. tidygraph inherits multiple algorithms from igraph, including edge-betweenness, leading eigenvector, fast-greedy, Louvain, walktrap, label propagation, InfoMap, spinglass and optimal.

g <- GAStech_graph %>%
  mutate(community = as.factor(
    group_edge_betweenness(
      weights = Weight,
      directed = TRUE))) %>%
  ggraph(layout = "fr") +
  geom_edge_link(
    aes(width = Weight),
    alpha = 0.2) +
  scale_edge_width(
    range = c(0.1, 5)) +
  geom_node_point(
    aes(colour = community))

g + theme_graph()

Things to learn from the code chunk above

group_edge_betweenness() returns a community ID per node, which is cast to factor so ggraph treats it as a categorical aesthetic. The algorithm progressively removes the edges with the highest betweenness until the graph fragments into communities.

The community network is further refined by drawing concave hulls around each detected community with geom_mark_hull() from ggforce.

g <- GAStech_graph %>%
  activate(nodes) %>%
  mutate(community = as.factor(
    group_optimal(weights = Weight)),
         betweenness_measure = centrality_betweenness()) %>%
  ggraph(layout = "fr") +
  geom_mark_hull(
    aes(x, y,
        group = community,
        fill = community),  
    alpha = 0.2,  
    expand = unit(0.3, "cm"),
    radius = unit(0.3, "cm")
  ) +
  geom_edge_link(aes(width = Weight),
                 alpha = 0.2) +
  scale_edge_width(range = c(0.1, 5)) +
  geom_node_point(aes(fill = Department,
                      size = betweenness_measure),
                      color = "black",
                      shape = 21)

g + theme_graph()

Things to learn from the code chunk above

geom_mark_hull() from ggforce draws a smoothed concave hull around each community. expand controls how much padding sits between the outermost nodes and the hull edge, and radius controls the corner smoothness. Combining community membership (hull fill) with department (node fill) reveals whether the structural communities match the formal organisation chart.

Reading the plot

When hull membership matches department colour, the formal structure drives communication. Hulls that span multiple departments flag informal collaboration that the org chart misses.

27.8 Building Interactive Network Graph with visNetwork

visNetwork wraps the vis.js JavaScript library. It accepts a nodes data frame with an id column and an edges data frame with from and to columns.

27.8.1 Data preparation

GAStech_edges_aggregated <- GAStech_edges %>%
  left_join(GAStech_nodes, by = c("sourceLabel" = "label")) %>%
  rename(from = id) %>%
  left_join(GAStech_nodes, by = c("targetLabel" = "label")) %>%
  rename(to = id) %>%
  filter(MainSubject == "Work related") %>%
  group_by(from, to) %>%
    summarise(weight = n()) %>%
  filter(from != to) %>%
  filter(weight > 1) %>%
  ungroup()

Things to learn from the code chunk above

The two left_join() calls translate sender and recipient labels into the numeric IDs that visNetwork expects. The renames to from and to are mandatory because those are the column names the function looks for.

27.8.2 Plotting the first interactive network graph

visNetwork(GAStech_nodes,
           GAStech_edges_aggregated)

Things to learn from the code chunk above

visNetwork() accepts the nodes and edges directly. The default layout is a physics simulation that runs in the browser, so the graph drifts into place after the page loads.

27.8.3 Working with layout

visNetwork(GAStech_nodes,
           GAStech_edges_aggregated) %>%
  visIgraphLayout(layout = "layout_with_fr")

Things to learn from the code chunk above

visIgraphLayout() pre-computes the layout using an igraph algorithm (layout_with_fr for Fruchterman-Reingold) and disables the browser-side physics. The visible benefit is a static, reproducible layout that loads instantly.

27.8.4 Working with visual attributes - Nodes

GAStech_nodes <- GAStech_nodes %>%
  rename(group = Department)

visNetwork(GAStech_nodes,
           GAStech_edges_aggregated) %>%
  visIgraphLayout(layout = "layout_with_fr") %>%
  visLegend() %>%
  visLayout(randomSeed = 123)

Things to learn from the code chunk above

visNetwork colours nodes automatically when there is a column named group, so renaming Department to group is enough to trigger colour-by-department. visLegend() adds a clickable legend, and visLayout(randomSeed = 123) makes the layout deterministic across reloads.

27.8.5 Working with visual attributes - Edges

visNetwork(GAStech_nodes,
           GAStech_edges_aggregated) %>%
  visIgraphLayout(layout = "layout_with_fr") %>%
  visEdges(arrows = "to",
           smooth = list(enabled = TRUE,
                         type = "curvedCW")) %>%
  visLegend() %>%
  visLayout(randomSeed = 123)

Things to learn from the code chunk above

arrows = "to" adds arrowheads at the destination end, restoring the directional information that the static ggraph examples did not encode. smooth = list(enabled = TRUE, type = "curvedCW") bends edges into clockwise curves so reciprocal pairs no longer superimpose.

27.8.6 Interactivity

visNetwork(GAStech_nodes,
           GAStech_edges_aggregated) %>%
  visIgraphLayout(layout = "layout_with_fr") %>%
  visOptions(highlightNearest = TRUE,
             nodesIdSelection = TRUE) %>%
  visLegend() %>%
  visLayout(randomSeed = 123)

Things to learn from the code chunk above

highlightNearest = TRUE dims everything except the clicked node and its direct neighbours, isolating one ego network at a time. nodesIdSelection = TRUE adds a dropdown for selecting a node by ID, which scales better than clicking when the network has many nodes.

Reading the plot

Click-to-highlight turns the static map into a single-node investigation tool. The viewer can pick any employee and immediately see who they communicate with, which is the same question a static facet plot answers at the cost of an entire panel.

27.9 Reference

Kam, Tin Seong. R for Visual Analytics. Singapore Management University, 2024.

🕹️ LEVEL COMPLETE 🕹️

★ ★ ★ ★ ★
CHAPTER 27 CLEARED!+1000 XP · ACHIEVEMENT UNLOCKED: Network Navigator 🕸️
Press any key to continue…

27 Chapter 27: Modelling, Visualising and Analysing Network Data with R

27.1 Overview

27.2 Getting Started

27.2.1 Installing and launching R packages

27.3 The Data

27.3.1 The edges data

27.3.2 The nodes data

27.3.3 Importing network data from files

27.3.4 Reviewing the imported data

27.3.5 Wrangling time

27.3.6 Reviewing the revised date fields

27.3.7 Wrangling attributes

27.3.8 Reviewing the revised edges file

27.4 Creating network objects using tidygraph

27.4.1 The tbl_graph object

27.4.2 The dplyr verbs in tidygraph

27.4.3 Using tbl_graph() to build tidygraph data model

27.4.4 Reviewing the tidygraph graph object

27.4.5 Interpreting the tidygraph graph object

27.4.6 Changing the active object

27.5 Plotting Static Network Graphs with ggraph package

27.5.1 Plotting a basic network graph

27.5.2 Changing the default network graph theme

27.5.3 Changing the colouring of the plot

27.5.4 Working with ggraph’s layouts

27.5.5 Fruchterman and Reingold layout

27.5.6 Modifying network nodes

27.5.7 Modifying edges

27.6 Creating facet graphs

27.6.1 Working with facet_edges()

27.6.2 Adjusting the legend position

27.6.3 A framed facet graph

27.6.4 Working with facet_nodes()

27.7 Network Metrics Analysis

27.7.1 Computing centrality indices

27.7.2 Visualising network metrics

27.7.3 Visualising Community

27.8 Building Interactive Network Graph with visNetwork

27.8.1 Data preparation

27.8.2 Plotting the first interactive network graph

27.8.3 Working with layout

27.8.4 Working with visual attributes - Nodes

27.8.5 Working with visual attributes - Edges

27.8.6 Interactivity

27.9 Reference

🕹️ LEVEL COMPLETE 🕹️

★ ★ ★ ★ ★

CHAPTER 27 CLEARED!