Comparison of image read performance of imager and magick packages.

imager_vs_magick-read_perf.R

# You will need to generate images of varying size using ImageMagic's convert utility on the shell.
# The shell code below will not work under Windows. Sorry! bash-only.
# Run the following shell code to generate some random images
# mkdir -p ./data/RandomNoise
# (cd ./data/RandomNoise && for i in `seq 100 100 4000`; do convert -size ${i}x${i} xc:   +noise Random   random-${i}_$i.png;done)

# You'll need the following R packages
# dplyr, purrr, magick, imager, magrittr, gridExtra, ggplot2, stringr

d <- './data/RandomNoise'

files <- list.files(d)
files.sizes <- purrr::map_dbl(files,function(x) file.size(sprintf("%s/%s",d,x)))

files.df <- data.frame(path=stringr::str_c(d,'/',files),name=files,size=files.sizes, stringsAsFactors = F) 
files.df %<>%
  dplyr::arrange(size)

# To format our X axis
format_si <- function(...) {
  # Format a vector of numeric values according
  # to the International System of Units.
  # http://en.wikipedia.org/wiki/SI_prefix
  #
  # Based on code by Ben Tupper
  # https://stat.ethz.ch/pipermail/r-help/2012-January/299804.html
  # Args:
  #   ...: Args passed to format()
  #
  # Returns:
  #   A function to format a vector of strings using
  #   SI prefix notation
  #
  
  function(x) {
    limits <- c(1e-24, 1e-21, 1e-18, 1e-15, 1e-12,
                1e-9,  1e-6,  1e-3,  1e0,   1e3,
                1e6,   1e9,   1e12,  1e15,  1e18,
                1e21,  1e24)
    prefix <- c("y",   "z",   "a",   "f",   "p",
                "n",   "µ",   "m",   " ",   "k",
                "M",   "G",   "T",   "P",   "E",
                "Z",   "Y")
    
    # Vector with array indices according to position in intervals
    i <- findInterval(abs(x), limits)
    
    # Set prefix to " " for very small values < 1e-24
    i <- ifelse(i==0, which(limits == 1e0), i)
    
    paste(format(round(x/limits[i], 1),
                 trim=TRUE, scientific=FALSE, ...),
          prefix[i])
  }
}

# magick performance

magick_image_read.times <- purrr::map(files.df$path, function(x) system.time(magick::image_read(x)))

magick_image_read.times.elapsed <- sapply(magick_image_read.times,function(x) x[[3]])
magick_image_read.times.system <- sapply(magick_image_read.times,function(x) x[[2]])
magick_image_read.times.user <- sapply(magick_image_read.times,function(x) x[[1]])
magick_image_read.times.df = data.frame(size=files.df$size, user=magick_image_read.times.user, 
                                        system=magick_image_read.times.system,
                                        elapsed=magick_image_read.times.elapsed,
                                        stringsAsFactors = F)
magick_image_read.times.df.tidy <- magick_image_read.times.df %>% tidyr::gather(key='metric',value = 'time',-size)

magick_image_read.plot <- ggplot(magick_image_read.times.df.tidy,aes(x=size,y=time,color=metric)) + geom_point() +
  geom_smooth( se=FALSE, alpha=0.4) + theme_minimal() +
  scale_x_continuous(labels=format_si())


# imager performance ----
imager_image_read.times <- purrr::map(files.df$path, function(x) system.time(imager::load.image(x)))

imager_image_read.times.elapsed <- sapply(imager_image_read.times,function(x) x[[3]])
imager_image_read.times.system <- sapply(imager_image_read.times,function(x) x[[2]])
imager_image_read.times.user <- sapply(imager_image_read.times,function(x) x[[1]])
imager_image_read.times.df = data.frame(size=files.df$size, user=imager_image_read.times.user, 
                                        system=imager_image_read.times.system,
                                        elapsed=imager_image_read.times.elapsed,
                                        stringsAsFactors = F)
imager_image_read.times.df.tidy <- imager_image_read.times.df %>% tidyr::gather(key='metric',value = 'time',-size)

imager_image_read.plot <- ggplot(imager_image_read.times.df.tidy,aes(x=size,y=time,color=metric)) + geom_point() +
  geom_smooth( se=FALSE, alpha=0.4) + theme_minimal() +
  scale_x_continuous(labels=format_si())
gridExtra::grid.arrange(magick_image_read.plot + labs(x='Image Size', y='Time (sec)', title='magick::image_read()', colour='Time'), 
                        imager_image_read.plot + labs(x='Image Size', y='Time (sec)', title='imager::load.image()', colour='Time'),
                        ncol=2)