memify (version 0.1.1)

memify-package: A Simple Framework to Construct and Maintain Functions That Keep State

Description

This package provides a simple way to construct and maintain versions of R functions that keep their state -- i.e. "remember" their explicitly specifed arguments from previous calls. If not overridden, the remembered values of these arguments are automatically reused as the defaults for subsequent calls.

This may be convenient when a function with a large argument list -- plotting functions are typical examples -- needs to be repeatedly invoked with only a few changes in the arguments at each invocation; or when interacting with a function to determine what parameter values give the most informative results. While there are certainly other ways to do this, the simplicity of this approach may make it a useful alternative.

Details

The package has only one key function: memify(). It takes one argument, f, a function or a function name as a name/symbol or character string. The result, almost always assigned to a different name, is a new version of f that will keep state. It has S3 class "memified" that extends the function's original class vector . The original version of f of course remains.

Some minor additional convenience functions, arglist(), arglist(x) <- value, and an update method, update.memified(), are also provided. See the examples here and in their repective Help pages for usage.

Examples

# NOT RUN {
mod <- function(x, b = 5) x %% b
mod.m <- memify(mod) ## or memify("mod")
mod.m(7) ## using default b = 5
mod.m(b = 3) ## using remembered x = 7
mod.m()  ## the same as previous

arglist(mod.m) ## list of all memified arguments
update (mod.m, b = 9) ## silently updates arglist
arglist(mod.m)

arglist(mod.m) <- list(x=11) ## replaces the arglist with a new list
arglist(mod.m)
## b is no longer memified, so mod's default = 5 is used:
mod.m()

## cleanup
rm(mod.m)
# }