Open In App

We use cookies to ensure you have the best browsing experience on our website. By using our site, you acknowledge that you have read and understood our Cookie Policy & Privacy Policy

networkD3 package in R

networkD3: Data-driven document Network, is an R package for creating network graphs which are used for 3-Dimensional visualizations of data as network graphs. In R Programming Language It was created using the HTML widget package in javascript by Christopher Gandrud et.al, Using (>=v0.99) this version of Rstudio the graph will appear on the viewer pane. Not only appear but also gives us a handy way of seeing and tweaking our network graphs. This package helps understand the network connections between nodes. Moreover, if you want to extract the graph from Rstudio you can export the graph to the clipboard or PNG/JPEG/etc.

Note: You are used to R’s 1-based indexing (i.e. counting in R starts from 1). However, networkD3 plots are created using JavaScript, which is 0-based indexing so, your data links will need to start from 0



In the ,networkD3 package there are many functions. Following are the popular network functions used by many.

INSTALLATION:

To use a package in R programming we have to install the package first. For installing the R package in R studio use the command install.packages(“name”). Follow the following steps to get the packages installed on your system.



install.packages('networkD3')

SIMPLE NETWORK

A simple network is a basic form of a Forced-directed network, we can use Force network to create a Simple network by providing appropriate data and customization options. For a simple form of a Force-directed network, we can use a simple network (). It works based on a “node-link” diagram and is connected by edges. It provides us with amazing animation with an intuitive way of observing data and also allows us to tweak the networks with cursors. The purpose of the simple network is to provide a good understanding of difficultly visualized data. It is used to represent simple connections like social media connections, biological connections(Genes, Proteins,..), etc, It contains few network connections compared to the force network since it’s the base variant of the force-directed network.

SYNTEX

simpleNetwork(Data, Source = 1, Target = 2, height = NULL, width = NULL, linkDistance = 50
,charge = -30, fontSize = 7, fontFamily = "serif",   linkColour = "#666", 
nodeColour = "#3182bd",opacity = 0.6, zoom = F)

ARGUMENTS:

Example
For Instance, we have a group of five friends named Sam, Sai, Robert, Tom, and Murphy. They are talking and sharing some information among themselves this simple conversation among those friends is represented as edges and the five friends are represented as nodes. This scenario of discussion among them can be visualized in a simple network.




# Load package
library(networkD3)
 
# Create data
X <- c("A","A","A","A",
       "B","C","B","D","C")
Y <- c("C","B","D","E",
       "K","J","G","H","I")
networkData <- data.frame(X,Y)
 
# Plot
simpleNetwork(networkData, fontFamily='fantasy', nodeColour = 'green') 





                        









Output:


NetworkD3




FORCE NETWORK
A  Force network is also known as node-link diagram though it is useful for visualizing networks with nodes and edges which are connected by forces. Force network provides wonderful animations for understanding the network’s data element and also allows us to tweak the data in a repulsive and attractive manner. For the complex form of a forced-directed network and to have more control over the appearance of the forced-directed network we can use force network (). It contains more network connections than a simple network since it’s the extension variant of the forced-directed network.
SYNTAX:


forceNetwork(Links, Nodes, Source, Target, Value, NodeID, Nodesize, Group,height = NULL,
width = NULL,   colourScale = JS("d3.scaleOrdinal(d3.schemeCategory20);"), fontSize = 7,
fontFamily = "serif", linkDistance = 50,   linkWidth = JS("function(d) 
{ return Math.sqrt(d.value); }"),
radiusCalculation = JS(" Math.sqrt(d.nodesize)+6"), 
charge = -30,linkColour = "#666", opacity = 0.6, zoom = FALSE, legend = FALSE,
arrows = FALSE, bounded = FALSE, opacityNoHover = 0,clickAction = NULL)
ARGUMENTS:


EXAMPLE
For instance consider there is a lakh of people living in a city with their friends, and families. The activity of contacting, sharing, and connecting among them can be represented as edges and the people, their friends, and their families are represented as nodes. The scenario of connection among a lack of people can be visualized using a force network.
















                                    



                                    




                                    




                                    


                                



















# Load data



data(MisLinks)




data(MisNodes)



 


# Plot



forceNetwork(Links = MisLinks, Nodes = MisNodes,




             Source = "source", Target = "target",




             Value = "value", NodeID = "name", 




             Group = "group", opacity = 1)




forceNetwork()



















                        






                        









Output:


NetworkD3


The ‘MisLinks’ and ‘MisNodes’ are an example of datasets in the ‘networkD3’ package that provides a simulated network of links (edges). MisLinks dataset contains columns such as ‘source’, ‘target’, and ‘value’ that describes the source nodes, target nodes, and the strength of the links. MisNodes dataset contains columns such as ‘name’, ‘group’, and ‘size’ that describes the node names, grouping, and sizes.


SANKEY NETWORK
A Sankey network is a type of visualization that shows the flows and movement of objects between nodes of different categories or groups. Sankey is useful for understanding the transition, distribution, or transformation of quantities or values within a system. It also consists of nodes and links. It is a complex form of network in the networkD3 package because of the massive network components. The link in the Sankey network is of different sizes according to the data flow through the link. The width of the links represents the magnitude or volume of the flow.
SYNTAX:
sankeyNetwork(Links, Nodes, Source, Target, Value, NodeID, NodeGroup = NodeID,
LinkGroup = NULL, units = "",   colourScale = JS("d3.scaleOrdinal(d3.schemeCategory20);"),
fontSize = 7,   fontFamily = NULL, nodeWidth = 15, nodePadding = 10, margin = NULL, 
height = NULL, width = NULL, iterations = 32, sinksRight = TRUE)
ARGUMENTS:


EXAMPLE:


Consider there is a number of industries, transportation, aviation, living areas, etc., that need resources like gas, oil, electricity, etc., for their existence and surveillance. Then the industries, transportation, gas, oil, etc., are represented as nodes of the network, and the connections among them can be represented as links(edges) of the network, If we require the amount of usage of the resources used by industries then we can use Sankey network.
















                                    



                                    




                                    




                                    


                                




















library(networkD3)



 


# Load energy projection data



URL <- paste0('https://cdn.rawgit.com/christophergandrud/networkD3/',




              'master/JSONdata/energy.json')




energy <- jsonlite::fromJSON(URL)



 


# Colour Nodes



sankeyNetwork(Links = energy$links, Nodes = energy$nodes, Source = 'source',




              Target = 'target', Value = 'value', NodeID = 'name',




              units = 'TWh', fontSize = 12, nodeWidth = 30)



 


# Colour links



energy$links$energy_type <- sub(' .*', '',




                                energy$nodes[energy$links$source + 1, 'name'])



 



sankeyNetwork(Links = energy$links, Nodes = energy$nodes, Source = 'source',




              Target = 'target', Value = 'value', NodeID = 'name',




              LinkGroup = 'energy_type', NodeGroup = NULL)



















                        






                        









Output:


NetworkD3


“jsonlite::fromJSON” is a function in “jsonlite” package in R that is used to convert JSON data in R objects that represents the JSON data. In the above program, jsonlite::fromJSON(URL) is a function in ‘jsonlite’ package that converts the JSON formatted data into R objects and stores it in the ‘energy’ variable.


RADIAL NETWORK:
A radial network also known as the is Reingold-Tilford Tree network, is a type of network visualization where nodes are arranged in a circular or radial layout. In a , radial network the root node is the center present at the centre of the network and the parent node present at is outside of the circular graph.  Radial network visualization is particularly useful when we want to display hierarchical relationships between nodes, such as organizational structures, family trees, or classification hierarchies.
SYNTAX:
radialNetwork(List, height = NULL, width = NULL, fontSize = 10, fontFamily = "serif",
linkColour = "#ccc", nodeColour = "#fff",  nodeStroke = "steelblue", textColour = "#111", 
opacity = 0.9, margin = NULL)
ARGUMENTS:


EXAMPLE:
For example, consider there is a group of employees working in an organization for the development of the firm. So, they require communication in a circulated manner for sharing of information among them. For this type of scenario, we can use a Radial network(tree network). An organization manager, employer, employee, HR, etc., can be represented as a node of the network and the sharing of information among them and their relationship can be represented as the links(edges).
















                                    



                                    




                                    




                                    


                                



















#Load library



library(networkD3)




Data <- paste0(




        "https://cdn.rawgit.com/christophergandrud/networkD3/",




        "master/JSONdata//flare.json")



 


## Convert to a list format



Flare <- jsonlite::fromJSON(Data, simplifyDataFrame = FALSE)



 


# Use a subset of data for more readable diagram


Flare$children = Flare$children[1:3]


 



radialNetwork(List = Flare, fontSize = 10, opacity = 0.9)



radialNetwork


















                        






                        









Output:




RADIAL NETWORK




DENDRO NETWORK
The Dendro network is a network visualization that uses a hierarchical type of visualization of data. It is particularly useful when visualizing `hierarchical clustering or tree structures. It contains the root node where the tree graph starts and the parent node where the tree graph ends. It can help for understanding complex information through network visualization.
SYNTAX:
dendroNetwork(hc, height=500, width=800, fontSize=10, linkColour="#ccc",nodeColour="#fff",
nodeStroke="steelblue",  textColour="#111", textOpacity=0.9,textRotate=NULL, opacity=0.9,
margins=NULL, linkType=c("elbow", "diagonal"),
treeOrientation=c("horizontal", "vertical"), zoom=FALSE)
ARGUMENTS:


EXAMPLE:
Consider there is a family consisting of family members and their ancestors, If a Researcher want to collect ancient data about a family like their ancestors related details and he/she wants to visualize those data in an understandable and structured form. In this type of scenario, a Researcher can choose Dendro Network. The family members, their parents, their grandparents, their great grandparents, etc., can be represented as nodes of the network, and the relationship links between them can be represented as edges.
















                                    



                                    




                                    




                                    


                                



















#Load library



library(networkD3)



 


#Creating data



hc <- hclust(dist(USArrests), method="ave")



 


#dendro_network



dendroNetwork(hc,textColour=c("red","green","orange")[cutree(hc,3)], height=600)



















                        






                        









Output:


DENDRO NETWORK




CHORD NETWORK
A Chord network is also known as a circular network. It is a type of network visualization that represents relationships and connections between entities or categories. It is particularly useful for showing the interactions and associations between different groups or entities. In a chord network, entities or groups are represented as arcs on a circle, and the connections or relationships are displayed as chords connecting the arcs. The width of the chords can be used to represent the strength, magnitude, or frequency of the connections.
SYNTAX:
chordNetwork(Data, height=500, width=500, initialOpacity=0.8, useTicks=0,colourScale=NULL,
padding=0.1, fontSize=14, fontFamily="sans-serif", labels=c(), labelDistance=30)
ARGUMENTS:


EXAMPLE:
In supply chain marketing, Chord diagrams can be used to visualize supply chain relationships, showing connections between suppliers, manufacturers, distributors, and customers. This can help identify dependencies, bottlenecks, and potential improvements in the supply chain.
















                                    



                                    




                                    




                                    


                                



















#Load the library



library(networkD3)



 


#Create data



hairColourData<-matrix(c(11975,1951,8010,1013, 




                         5871,10048,16145,990, 




                         8916,2060,8090,940, 




                         2868,6171,8045,6907), 




                         nrow=4) 



 


#Create chord diagram



chordNetwork(Data=hairColourData, 




             width=500, 




             height=500, 




             colourScale=c("#000000", "#FFDD89", "#957244", "#F26223"), 




             labels=c("red","brown","blond","gray"))



















                        






                        









Output:


chord Network




SAVING NETWORK GRAPH AS HTML FILE
If you want to save your network in an HTML file you can use saveNetwork it after the network implementation. The function saveNetwork is from the library ‘magrittr’. Using the below R code in Rstudio we can save the network of any type in our system as HTML, JPEG, or PNG file














                                    



                                    




                                    




                                    


                                




















library(magrittr)



 



simpleNetwork(networkData) %>%




saveNetwork(file = 'Net1.html')



















                        






                        








	

        Article Tags : 
        

            R Language        


	


      

      

          
Recommended Articles

      

      

          
          
1. How to add Axis labels using networkD3 in R

          
2. tidyr Package in R Programming

          
3. Deploying a basic Streamlit app using Shiny package in R

          
4. How to Plot in 3D clusters using plotly package

          
5. psych Package in R

          
6. Density Plots Using Lattice Package in R

          
7. Car package in R

          
8. Desctools Package in R

          
9. What is the Glmnet package in R?

          
10. Purrr Package in R Programming

          
11. Rank variable by group using Dplyr package in R

          
12. Data Visualization using GoogleVis Package

          
13. Introduction to emojifont package in R

          
14. Create a Choropleth Map by using Plotly Package in R

          
15. Esquisse Package in R Programming

          
16. Variable importance plot using random forest package in R

          
17. Data Visualization using ggvis Package in R

          
18. How to Install a Package in R ?

          
19. How to disable Messages when Loading a Package in R ?

          
20. Cross validation in R without caret package

          
21. Draw Multiple Overlaid Histograms with ggplot2 Package in R

          
22. How to create a scatter plot using lattice package in R?

          
23. How to create a boxplots using lattice package in R?

          
24. Working With Different Versions of an R Package

          
25. Union() & union_all() functions in Dplyr package in R