How to add Axis labels using networkD3 in R

networkD3 is an R package used for creating a D3 (Data-Driven Documents) Network Graph. netwrorkD3 is constructed using the htmlwidget package. As the name said network, this graph can be constructed in the shape of a node and edge data frame. then it will perform a physics simulation to decide the most advantageous diagram in three dimensions. It is used to create a D3 Javascript network, tree, dendrogram, Sankey graph, etc.

In the R package, there are already many visualization libraries present. For example ggplot, Plotly, igraph, etc. But not all libraries are used to create network-type graphs and it is also not in 3D plots. That is the main objective behind why it is made.

Syntax: sankeyNetwork( Links, Nodes, Source, Target, Value, NodeID, NodeGroup = NodeID,                   LinkGroup = NULL, units = "", colourScale = NULL, fontSize = 7, fontFamily = NULL, nodeWidth = 15, nodePadding = 10, margin = NULL, height = NULL, width = NULL, iterations = 32, sinksRight = TRUE )

Parameters:

• Links: dataframe represent link between source and target (node).
• Nodes: dataframe contain node properties
• Source: character string of source node
• Target: character string of target node
• Value: character string of distance value between nodes
• NodeID: node name
• NodeGroup: character string of the nodeIDs in nodes
• LinkGroup: character string for link between groups
• Units: character string to specify physical unit(value)
• colourscale: specify color scale
• nodeWidth: width of node
• nodePadding: padding between node
• sinksRight: Boolean. If it is True, last node moves to right border of graph

Creating a simple sankeyNetwork graph with Title, Subtitle, and Labels

Step: 1 Import the necessary library in Rstudio.

library(dplyr)
library(networkD3)
library(htmlwidgets)
library(manipulateWidget)
library(htmltools)

Step: 2 Create network data in a source-target data frame format.

• Here, we are linking A->D, B->E, and C->F nodes and assigning 3 values on each link. final table will look like myDataframe table image.
• The node table is containing every node entry uniquely.
• IDsource and IDtarget is a string character form of the source and the target node.
• match function will match take two-column and create a matching table from 1. But we are subtracting by '-1' so that it can start with 0th character like for A node it is 0, B node it is 1 and so on.

# Create values
Source_Node <- c("A","B","C","A","B",
                 "C","A","B","C")
Target_category <- c("D","E","F","D",
                     "E","F","D","E","F")
Frequency <- c(1,2,3,4,5,6,7,8,9)
myDataframe <- data.frame(Source_Node,
                          Target_category,Frequency)

# Node table
nodes <- data.frame(name=c(as.character(myDataframe$Source_Node), 
                           as.character(myDataframe$Target_category))
                    %>% unique())

# IDsource and IDtarget string character format
myDataframe$IDsource=match(myDataframe$Source_Node, 
                           nodes$name)-1 
myDataframe$IDtarget=match(myDataframe$Target_category,
                           nodes$name)-1

Output:

Step: 3 Create a simple Sankey network chart.

As per the above sankeyNetwork syntax, we are assigning some parameters to generate our output.

graph1 <- sankeyNetwork(Links = myDataframe,  
                        Nodes = nodes, 
                   Source = "IDsource", 
                        Target = "IDtarget", 
                   Value = "Frequency",
                        NodeID = "name",  
                   sinksRight=FALSE)

# view graph
graph1

Output:

Step: 4 Add Title and Subtitle in graph1.

• To assign a title, we can use the prependContent function from the htmlwidgets tool. 
• Create two h1 and h2 tags with text description and style.
• prepend these tags in our above graph1 by putting all these tags together in the prependContent function.

graph2 <- htmlwidgets::prependContent(graph1, 
                                      htmltools::tags$h1(
                                        "GeeksForGeeks Title",
                                      style="text-align:center; 
color:#007500; "),
                                      htmltools::tags$h2(
                                        "GeeksForGeeks Subtitle",
                                      style="color:#00D100; 
text-align:center;"))

# view graph
graph2

Output:

Step: 5 Add left and right labels in networkGraph.

• create left-tag as leftTx and right-tag as rightTx.
• Now, to combine leftTx, rightTx and graph2, use combineWidget from manipulatewidget library.
• combineWidget take, graph, title, left-column, right-column, and row-number as parameters and combine all these instances into one.

Syntax: combineWidgets( list = NULL, nrow = NULL, ncol = NULL, title = NULL, rowsize = 1, colsize = 1, byrow = TRUE, titleCSS = "", header = NULL, footer = NULL, leftCol = NULL, rightCol = NULL, width = NULL, height = NULL )

# create left-right label 
leftTx = tags$div( 
         style="max-width: 30vw; height: 100%; 
                 display: flex; align-items: center; 
                 justify-content: center;", 
         tags$p("This is left side label"))

rightTx = tags$p("This is right side label",
                   style="max-width:30vw")

# final output with label. combineWidget
# can be use by manipulateWidget.
graph3 <- combineWidgets(graph2,
                      title = tags$h1("With left-right label",
                      style="text-align:center;color:#007500;"),
                      leftCol = leftTx,
                      rightCol = rightTx,
                      nrow = 1)

# view graph
graph3

Output: