Batch save plots to multiple formats.

This is a formalized version of a function I often have in utils scripts for different projects. It works well when paired with purrr::map, taking a list of plots and a list of parameters for each (width, height, filename), and saving each to both PDF (good as a vector format) and PNG (good as a decent resolution bitmap format) files using ggplot2::ggsave. By default, this will add a logo to the bottom of the plot using add_logo—look at the documentation to see exactly how to pass a logo. Note that some devices might act differently: for example, SVG devices might have some quirks.

Usage

write_plot(
  plot,
  filename,
  width = 7,
  height = 4.5,
  add_logo = TRUE,
  use_abs_logo = TRUE,
  logo_abs = 0.15,
  logo_scale = 0.05,
  logo_img = NULL,
  logo_pos = "right",
  place_inside = FALSE,
  dir = NULL,
  separate_dirs = TRUE,
  devs = list(pdf = grDevices::cairo_pdf, png = ragg::agg_png),
  verbose = TRUE,
  ...
)

Arguments

plot: A ggplot object to save
filename: String, the filename without extension, e.g. "cool_plot" rather than "cool_plot.png", since extensions will be added based on the devs argument.
width: Numeric, width of saved plot in inches. Default: 7
height: Numeric, height of saved plot in inches. Doesn't include any additional height for the logo. Default: 4.5
add_logo: Logical: if TRUE (default), a logo will be added to each plot using add_logo.
use_abs_logo: Logical, whether the logo size is being supplied as absolute (if TRUE) or relative to the plot height (if FALSE). If TRUE (default), the logo will be the height given in logo_abs in inches; otherwise, the logo will be sized as the ratio to the plot height as given in logo_scale. For batch writing plots, using an absolute size is probably best so the logo is the same size on all charts, regardless of each chart's height.
logo_abs: The absolute height for the logo, in inches. Ignored if use_abs_logo = FALSE. Default: 0.15
logo_scale: The ratio of the logo's height to the plot's height. Ignored if use_abs_logo = TRUE. Default: 0.05
logo_img: If NULL (the default), the logo will be the image saved at system.file("extdata/logo.svg", package = "stylehaven"), which is a copy of DataHaven's logo. See add_logo for other ways to create and supply a logo.
logo_pos: String, which side of the plot the logo should be placed on. Default: 'right'
place_inside: Logical, whether the logo should be placed on the inside of the chart or the outside (default). Note that this is the opposite of the default in add_logo—often placing the logo inside the chart works well, but for batch writing it's probably safer to either default to outside placement, or use a parameter for this alongside the width & height.
dir: Directory in which to write files. If NULL, the default, this will just be the working directory, as is the case with ggplot2::ggsave.
separate_dirs: Logical, whether to save files into separate directories based on file type. If TRUE (default), the file "cool_plot.png" will be written to "dir/png/cool_plot.png"; otherwise, the file will be written just in dir. If the necessary directories don't already exist, they'll be created.
devs: A named list, where the names are the extensions to use for output files and the values specify graphics devices to use, either as strings or functions (or a mix). See the device argument of ggplot2::ggsave() for examples. The defaults, c(pdf = cairo_pdf, png = ragg::agg_png), are the ones that I've found to work well for writing out ggplots, particularly high-resolution ones with custom fonts.
verbose: Logical, whether to print the path to each file after it's written (defaults TRUE).
...: Additional arguments to pass on to ggplot2::ggsave, which in turn passes them to graphics devices.

Value

Returns nothing. If verbose = TRUE, sends a message to the console.

Examples

if (FALSE) {
  library(ggplot2)
  plot1 <- ggplot(mtcars, aes(x = hp)) +
    geom_density()
  write_plot(plot1, "hp_density")
  write_plot(plot1, "hp_density", devs = list(png = "png", svg = svg))
  ggsave("svg/plot.svg", plot1, device = svglite::svglite)

  # how I usually use this for a whole document of plots
  plots <- list()
  plots[["mpg_histogram"]] <- ggplot(mtcars, aes(x = mpg)) +
    geom_histogram()
  plots[["hp_vs_mpg"]] <- ggplot(mtcars, aes(x = hp, y = mpg)) +
    geom_point()
  plot_params <- list(
    mpg_histogram = list(w = 7, h = 4),
    hp_vs_mpg = list(w = 5, h = 5)
  )
  purrr::imap(plots, function(plt, id) {
    # using plot ID as the filename
    params <- plot_params[[id]]
    write_plot(plt, filename = id,
               width = params$w,
               height = params$h,
               logo_abs = 0.2)
  })
}

Usage

Arguments

Value

See also

Examples