remake.yml

## All R files that are used must be listed here:
sources:
  - code.R

## Most of a remakefile will be targets.  They can be listed in any
## order; whatever makes most sense.  Files can be split up and
## included into the main file if you need to impose more structure.
targets:
  ## 'all' is a special target that will be used as the default
  ## target by 'remake'.  It doesn't actually make anything but
  ## "depends" on other targets.
  all:
    depends: plot.pdf

  ## 'target_name' is a special variable here: it will pass 'data.csv'
  ## through to the function 'download_data()' so that the data are
  ## stored in the correct place.
  data.csv:
    command: download_data(target_name)

  ## 'processed' will be stored as an R object because it looks like
  ## an object name rather than a file name (file names have
  ## extensions or slashes in them).  It will be saved across
  ## sessions.  Because it uses "data.csv", remake will work out that
  ## this target depends on "data.csv".
  processed:
    command: process_data("data.csv")

  ## Specifying 'plot: true' here tells remake that you are creating a
  ## plot.  It will sort out opening/closing the pdf device for you.
  ## Because 'myplot' uses 'processed', remake will work out that this
  ## target requires the processed target to be made first.
  plot.pdf:
    command: myplot(processed)
    plot: true

  ## Like plot targets, there are also special knitr targets.  This
  ## will arrange for "processed" to be available within the knit
  ## function, along with all functions defined in "code.R".  This
  ## automatically depends on the file report.Rmd, and will be updated
  ## when that file changes or when "processed" changes.
  report.md:
    depends: processed
    knitr: true