I've created a Sankey diagram in R, using the networkD3 package, that I'd like to save as a static image, using code instead of clicking on 'Export' --> 'Save as Image...'.
The current code I've tried (using this Sankey diagram as an example) is:
library(networkD3)
URL <- paste0(
"https://cdn.rawgit.com/christophergandrud/networkD3/",
"master/JSONdata/energy.json")
Energy <- jsonlite::fromJSON(URL)
# Plot
jpeg( filename = "Sankey.jpg", width = 4000, height = 4000)
sankeyNetwork(Links = Energy$links, Nodes = Energy$nodes, Source = "source",
Target = "target", Value = "value", NodeID = "name",
units = "TWh", fontSize = 12, nodeWidth = 30)
dev.off()
All I'm getting though is a blank white box when I open the image though.
Simplest working solution I've found so far is:
Install PhantomJS. For example using Homebrew for OSX - brew install phantomjs
Install rbokeh - install.packages("rbokeh")
Then:
library(rbokeh)
sn <- sankeyNetwork(Links = Energy$links, Nodes = Energy$nodes, Source = "source",
Target = "target", Value = "value", NodeID = "name",
units = "TWh", fontSize = 12, nodeWidth = 30)
widget2png(sn, "sankey.png")
The result doesn't look great, but this might serve as a starting point for research and improvements.
EDIT: here's another potential solution using the webshot package.
Related
I have created a interactive visualisation using the following code:
library(networkD3)
nodes = data.frame("name" = c("node1", "node2","node3", "node4", "node5", "node6", "node7"))
links = as.data.frame(matrix(c(
0,1,7937,
0,2,6990,
0,3,2483,
1,4,2120,
2,4,666,
3,4,282,
1,5,4583,
2,5,5657,
3,5,731,
1,6,1234,
2,6,756,
3,6,1470), byrow = TRUE, ncol = 3))
names(links) = c("source", "target", "value")
sankey <- sankeyNetwork(Links = links, Nodes = nodes,
Source = "source", Target = "target",
Value = "value", NodeID = "name",
fontSize= 12, nodeWidth = 15)'
This is my first time using the networkD3 package (or any interactive package for that matter) and from playing around I found that to keep it interactive it has to be published as a webpage (or is there another way??) but looking through the documentation for the package I can't see a way to add a title or a caption / comments. I want to share this piece of work round so need to explain what each level means on the published webpage ideally
There is no feature built-in to networkD3 to add titles or captions, but you can use functions in the htmlwidgets package to prepend or append content to an htmlwidget. There are numerous options, but for example....
library(htmlwidgets)
library(htmltools)
sankey <- htmlwidgets::prependContent(sankey, htmltools::tags$h1("Title"))
sankey <- htmlwidgets::appendContent(sankey, htmltools::tags$p("Caption"))
Responding to the comment, "I ended up using this to add a title, but it keeps pushing my viz down and cutting the bottom off. This stays off even when saving as a webpage. Is there anyway I can stop this from happening?"
I tried the suggested reply of adding sankey$sizingPolicy$viewer$fill <- FALSE, however, it made my sankey smaller than I wanted it. I found out that you can adjust the width and height of the Sankey prior to adding the HTML widget by adding width=(desired width) and height=(desired height) and this creates the space to then add in the title and the comment, as suggested by CJ Yetman.
library(networkD3)
library(htmlwidgets)
library(htmltools)
nodes = data.frame("name" = c("node1", "node2","node3", "node4", "node5", "node6", "node7"))
links = as.data.frame(matrix(c(
0,1,7937,
0,2,6990,
0,3,2483,
1,4,2120,
2,4,666,
3,4,282,
1,5,4583,
2,5,5657,
3,5,731,
1,6,1234,
2,6,756,
3,6,1470), byrow = TRUE, ncol = 3))
names(links) = c("source", "target", "value")
sankey <- sankeyNetwork(Links = links, Nodes = nodes,
Source = "source", Target = "target",
Value = "value", NodeID = "name",
fontSize= 12, nodeWidth = 15,
width= 900, height=600)
sankey <- htmlwidgets::prependContent(sankey, htmltools::tags$h1("Title"))
sankey <- htmlwidgets::appendContent(sankey, htmltools::tags$p("Caption"))
sankey
I am using networkD3 to create sankey diagrams in R. I am using something like below to generate the diagrams:
sankeyNetwork(Links = data$links, Nodes = data$nodes,
Source = "source", Target = "target", Value = "value", NodeID = "name",
units = "Questions", fontSize = 12, nodeWidth = 10,
colourScale = "d3.scale.category10()",
LinkGroup = "type", #<--assign colors to links
width = 1000, height = 600)
I am using LinkGroup = "type" to color the links, where the type is basically the source node's name. This makes the link color same as source node color. This is working pretty well.
I was wondering if there is anyway a gradient of color can be created from sources to targets for the associated links?
Is there a way to use a node as a link to an external website using the function forceNetwork() in the networkD3 package in r? I was thinking maybe modifying the clickAction?
Example data:
library(networkD3)
data(MisLinks)
data(MisNodes)
# Create a random URL in the nodes dataset
MisNodes$URL <- paste0("http://www.RANDOMLINK_", sample(1:100, NROW(MisNodes)), ".com")
head(MisNodes)
MyClickScript <- 'alert(d.index)'
forceNetwork(Links = MisLinks, Nodes = MisNodes,
Source = "source", Target = "target",
Value = "value", NodeID = "name",
Group = "group", opacity = 0.8,
clickAction = MyClickScript)
Desired outcome: When a user clicks on a node, a new tab will open (e.g. window.open) pointing to the associated URL for the node - How can I get clickAction to point to MisNodes$URL[d.index]?
networkD3 design does not make this easy. Here is one way to answer. I'll try to comment inline to explain what we are doing in each step.
library(networkD3)
# example from ?forceNetwork
data(MisLinks)
data(MisNodes)
# Create graph
fn <- forceNetwork(
Links = MisLinks, Nodes = MisNodes, Source = "source",
Target = "target", Value = "value", NodeID = "name",
Group = "group", opacity = 0.4, zoom = TRUE
)
# let's look at our forceNetwork
# nodes are provided to JavaScript
# in a nodes data.frame
str(fn$x$nodes)
# make up some links to demonstrate
# how we can add them to our nodes df
fn$x$nodes$hyperlink <- paste0(
'http://en.wikipedia.org/wiki/Special:Search?search=',
MisNodes$name
)
# then with our hyperlinks in our data
# we can define a click action to open
# the hyperlink for each node in a new window
fn$x$options$clickAction = 'window.open(d.hyperlink)'
fn
I am trying to plot Sankey diagrams using sankeyNetwork() in networkD3 package.
sankeyNetwork(Links = Flow_data, Nodes = Nodes_data,
Source = "Source_ID", Target = "Target",
Value = "value", NodeID = "Nodes_name",
width = 1000, height=600, fontsize = 16, nodeWidth = 50,
colourScale = "d3.scale.category20c()")
The visualization works great but I would like to change the colour range to an individual range. Is there any chance to change the colours of the SankeyNetwork? I need a range of only e.g. 3 colours which I can set by myself (not the predefined colourScales of d3.scale).
You can config:
sankeyNetwork(Links = Flow_data, Nodes = Nodes_data,
Source = "Source_ID", Target = "Target",
Value = "value", NodeID = "Nodes_name",
width = 1000, height=600, fontsize = 16, nodeWidth = 50,
colourScale = "d3.scale.category20c()") <==== Categorical color
UPDATE
Newer version:
d3.scale.ordinal().range(["#7d3945","#e0677b", "#244457"])
now works if changed to:
d3.scaleOrdinal().range(["#7d3945","#e0677b", "#244457"])
Thanks #Peter Ellis
UPDATE
Is there any way to set transparency when using custom colours?
"#AARRGGBB" doesn't seem to work
You can make a selectAll("your_class").style("opacity",0.5), Take a look to this: stackoverflow.com/questions/6042550/… for style attribute options. And CSS3 has a fully standardized solution: "fill="rgba(124,240,10,0.5)"
For color references, look here: http://bl.ocks.org/aaizemberg/78bd3dade9593896a59d
and here: https://github.com/mbostock/d3/wiki/Ordinal-Scales#categorical-colors
I am using the sankeyNetwork function in the networkD3 package in R using as an example the code found here. However, all I get is a blank screen. The diagram is supposed to show the flow of infections between age groups (by gender). My code is as below:
library(RCurl)
library(networkD3)
edges <- read.csv(curl("https://raw.githubusercontent.com/kilimba/data/master/infection_flows.csv"),stringsAsFactors = FALSE )
nodes = data.frame(ID = unique(c(edges$Source, edges$Target)))
nodes$indx =0
for (i in 1:nrow(nodes)){
nodes[i,]["indx"] = i - 1
}
edges2 <- merge(edges,nodes,by.x = "Source",by.y = "ID")
edges2$Source <-NULL
names(edges2) <- c("target","value","source")
edges2 <- merge(edges2,nodes,by.x = "target",by.y = "ID")
edges2$target <- NULL
names(edges2) <- c("value","source","target")
nodes$indx <- NULL
# Plot
sankeyNetwork(Links = edges2, Nodes = nodes,
Source = "source", Target = "target",
Value = "value", NodeID = "ID",
width = 700, fontsize = 12, nodeWidth = 30)
Are you sure there are no errors printed in your R console?
This works for me with two small modifications:
Load the curl package as well at the beginning
library("curl")
The fontsize parameter apparently does not work and should be removed.
# Plot
sankeyNetwork(Links = edges2, Nodes = nodes,
Source = "source", Target = "target",
Value = "value", NodeID = "ID",
width = 700, #fontsize = 12,
nodeWidth = 30)
Adjusting fontsize does work, but your argument is missing a capitalization: fontSize
sankeyNetwork(Links = edges2, Nodes = nodes,
Source = "source", Target = "target",
Value = "value", NodeID = "ID",
width = 700, fontSize = 12,
nodeWidth = 30)
you do not need RCurl, read.csv is able to read directly from a URL
it's probably safer to use the stringsAsFactors = FALSE option when creating the nodes data.frame
as others have pointed out, you must make sure that the source and target variables in the links data are numeric, and that they are zero-indexed
as others have pointed out, the font size parameter is properly named fontSize
I have provided a more direct way of creating the links data with numeric indexes of the nodes in the nodes data.frame
library(networkD3)
edges <- read.csv("https://raw.githubusercontent.com/kilimba/data/master/infection_flows.csv",stringsAsFactors = FALSE)
nodes = data.frame(ID = unique(c(edges$Source, edges$Target)), stringsAsFactors = FALSE)
edges$Source <- match(edges$Source, nodes$ID) - 1
edges$Target <- match(edges$Target, nodes$ID) - 1
sankeyNetwork(Links = edges, Nodes = nodes,
Source = "Source", Target = "Target",
Value = "Value", NodeID = "ID",
width = 700, fontSize = 12, nodeWidth = 30)
I solved it for me by making sure that source, target and values were all numeric.
For example:
Energy$links$value <- as.numeric(Energy$links$value)