PackageNetworks.R

#####
### The code below is written by Benjamin Lind under CC: BY-NC-SA 3.0
### If you use or develop this code, please provide proper attribution.
### You may contact me at lind.benjamin//at//gmail//d0t//com
### You may also find me @benjaminlind on Twitter
#####

#####
### See: http://blog.revolutionanalytics.com/2014/07/dependencies-of-popular-r-packages.html
### To run this script, enter: source("http://pastebin.com/raw.php?i=2Fa2Uz0N")
#####

library(igraph)
library(compiler)

ap <- available.packages()
ap <- ap[, c("Package", "Depends", "Suggests")]

pack2el <- function(packname, packmat, packrel){
  # _packname_ is the package name, a character
  # _packmat_ is the matrix produced by available.packages()
  # _packrel_ is the relationship between packages. Should be either "Depends" or "Suggests"
  ind <- which(packmat[, "Package"] == packname)
  alters <- packmat[ind, packrel]
  # Clean up the input
  alters <- gsub(" ", "", alters, fixed = TRUE)
  alters <- gsub("\n", "", alters, fixed = TRUE)
  alters <- gsub("\t", "", alters, fixed = TRUE)
  alters <- gsub("\r", "", alters, fixed = TRUE)
  alters <- unlist(strsplit(alters, ",", fixed = TRUE))
  # Drop R version dependency
  rvers <- grep("R(>", alters, fixed = TRUE)
  if (length(rvers) > 0)
    alters <- alters[-rvers]
  # Drop the version numbers of packages
  alters <- sapply(alters, function(x) return(strsplit(x, "(", fixed = TRUE)[[1]][1]))
  if (length(alters) == 0)
    alters <- NA
  outdfnrow <- length(alters)
  ego <- rep(packname, times = outdfnrow)
  if (packrel == "Depends")
    outdf <- data.frame(Sender = alters, Receiver = ego, stringsAsFactors = FALSE, row.names = 1:outdfnrow)
  else if (packrel == "Suggests")
    outdf <- data.frame(Sender = ego, Receiver = alters, stringsAsFactors = FALSE, row.names = 1:outdfnrow)
  return (outdf)
  }
pack2el <- cmpfun(pack2el)

depel <- do.call(rbind, lapply(ap[, "Package"], pack2el, packmat = ap, packrel = "Depends"))
depel <- depel[which(is.na(depel[, "Sender"]) == FALSE), ]
rownames(depel) <- 1:nrow(depel)

sugel <- do.call(rbind, lapply(ap[,"Package"], pack2el, packmat = ap, packrel = "Suggests"))
sugel <- sugel[which(is.na(sugel[, "Receiver"]) == FALSE), ]
rownames(sugel) <- 1:nrow(sugel)

packnames <- sort(unique(c(ap[,"Package"], depel[, "Sender"], sugel[, "Receiver"])))

dep.ig <- graph.data.frame(depel, TRUE, packnames)
sug.ig <- graph.data.frame(sugel, TRUE, packnames)

rm(ap, pack2el, depel, sugel, packnames)