RSiteCatalyst Sankey Diagram - Single Page to Multiple Pages

rsitecatalyst-sankey-one-page.R

library("RSiteCatalyst")
library("d3Network")

#### Authentication
SCAuth("key", "secret")

#### Get Pathing data: Single page, then ::anything:: pattern
pathpattern <- c("http://randyzwitch.com/big-data-hadoop-amazon-ec2-cloudera-part-1", "::anything::")
next_page <- QueuePathing("zwitchdev",
                          "2014-01-01",
                          "2014-08-31",
                          metric="pageviews",
                          element="page",
                          pathpattern,
                          top = 50000)

#Optional step: Cleaning my pagename URLs to remove to domain for clarity
next_page$step.1 <- sub("http://randyzwitch.com/","", 
                        next_page$step.1, ignore.case = TRUE)
next_page$step.2 <- sub("http://randyzwitch.com/","", 
                        next_page$step.2, ignore.case = TRUE)

#Get unique values of page name to create nodes df
#Create an index value, starting at 0
nodes <- as.data.frame(unique(c(next_page$step.1, next_page$step.2)))
names(nodes) <- "name"
nodes$nodevalue <- as.numeric(row.names(nodes)) - 1

#Convert string to numeric nodeid
links <- merge(next_page, nodes, by.x="step.1", by.y="name")
names(links) <- c("step.1", "step.2", "value", "source")

links <- merge(links, nodes, by.x="step.2", by.y="name")
names(links) <- c("step.1", "step.2", "value", "source", "target")

#Create next page Sankey chart
d3output = "C:/Users/rzwitc200/Desktop/sankey.html"
d3Sankey(Links = links, Nodes = nodes, Source = "source",
         Target = "target", Value = "value", NodeID = "name",
         fontsize = 12, nodeWidth = 100, file = d3output, width = 750, height = 600)