Script for loading and processing data on ggplot2 speed

ggplot2_speed_data.r

require(RCurl)
require(plyr)

mytsv <- getURL("https://docs.google.com/spreadsheet/pub?key=0AhQ4CGoYYLyvdDd6b2Q5MG9ic2NoQWRPT3dLT0FyUkE&output=txt", ssl.verifypeer = FALSE)  

raw <- read.delim(textConnection(mytsv), quote = "", stringsAsFactors = FALSE)

# Process the stuff from dput() on each row
parserow <- function(row) {
  d <- eval(parse(text = row[[2]]))

  data.frame(
    arch = d$ver$arch,
    os = d$ver$os,
    rver = paste(d$ver$major, d$ver$minor, sep = "."),
    time_build = d$time$user.self[2],
    time_render = d$time$user.self[3],
    time_draw = d$time$user.self[4],
    time_total = d$time$user.self[5]
  )
}

# Run parserow on each row and add to the raw data frame
dat <- adply(raw, 1, parserow)

# Drop the dput column
dat <- dat[,-2]

dat <- rename(dat, c(
  CPU.type="cpu",
  Number.of.cores="cores",
  Clock.speed="clock",
  Other.CPU.information="cpu.other",
  Short.description.of.computer="desc",
  Computer.Type="type",
  Other.notes="other"))

# Shorten description
dat$desc <- substr(dat$desc, 1, 16)

# Keep just some of the columns
d <- dat[,c("arch", "os", "cpu", "cores", "clock", "time_total", "rver", "type", "desc")]

# Sort by time
d <- arrange(d, time_total)

d