In crowding/nse: Tools for Lazy and Non-Standard Evaluation

These are the most common (and least problematic) scenarios for where the R user normally encounters or needs metaprogramming:

• Present to the user the same expressions that appear in the R code. For example, automatically generating plot labels.

• Saving information about how a function was called, so as to re-call it later. For example, model objects that allow derived models and application of different data.

• Forwarding your "raw" arguments to another function, which itself uses metaprogramming.

• For example, if you want to write a wrapper function around plot().

line.plot <- function(x, y, ...) {
  plot(x, y, ...)
}

line.plot <- function(x, y, ...) {
  args <- get_call()

  args <- c(args(x=x, y=y), dots(...))

  plot %()% args
}

line.plot <- function(x, y, ...) {

}

crowding/nse documentation built on Jan. 5, 2024, 12:14 a.m.