R - Function

About

Functions are stored as R objects with the class function

Functions can be:

used as arguments of other functions
nested

Articles Related

Syntax

function( arglist ) expr
return(value)

#of

f <- function(arglist) {
   ## Body
}

where:

arglist is a list of argument including the three dot argument.

Ex: arg1, arg2 = 1, arg3 = 'Nico', arg4 = NULL, …

Argument

Arguments mapping

Arguments can be matched positionally or by name.

Arguments mappings works as follow:

an exact match on the name occurs. The matched argument are removed from the argument list.
a partial match on the name occurs. The matched argument are removed from the argument list.
a positional match occurs on the position (the remaining unnamed arguments taking their position number from the remaining argument list)

All the below function calls are equivalent:

mean(x=data,na.rm=FALSE)
mean(na.rm=FALSE,x=data)
mean(data,na.rm=FALSE)
mean(na.rm=FALSE,data)

Args Function

To get the arguments of a function you can use the args function:

args(data.frame)

function (..., row.names = NULL, check.rows = FALSE, check.names = TRUE, 
    stringsAsFactors = default.stringsAsFactors()) 
NULL

Lazy evaluation

Arguments to functions are evaluated lazily (ie only when needed).

Argument is not used in the function

f = function(x, y) {
     return(x*2)
}
f(2)

[1] 4

Argument is not called

f = function(x, y) {
     print(x)
     print(y)
}
f(2)

[1] 2
Error in print(y) : argument "y" is missing, with no default

Notice that the error occurs only when the y argument is needed (ie interpreted)

The … argument

The … argument indicates a variable number of arguments.

This special argument is used :

when extending (wrapping) an existing function

myPlot = function(x, y, type = "l", ...) {
        plot(x, y, type = type, ...)
}

when the number of arguments cannot be known in advance.
by generic functions

Any arguments that appear after it must be named explicitly and cannot be partially matched.

Return Value

The return value of a function is:

returned by the return function
of by the last expression in the body function to be evaluated.

f = function(x, y) {
     return(x*2)
}

is equivalent to:

f = function(x, y) {
     x*2
}

is equivalent to:

f = function(x, y) {
     return(x*2)
     x*4 # This statement will be skipped
}

Environment

Typically, a function is defined in the global environment, so that the values of free variables are just found in the user’s workspace

In this case the environment in which a function is defined is the body of another function! ????

How to

Get help on function

Just use the question mark ? to obtain the documentation on a function

?data.frame

Get the parameters of a function

Use the str function

Example:

str(vector)
function (mode = "logical", length = 0L)

where:

function indicates that vector is a function
(mode = “logical”, length = 0L) are the parameters of the function.

Get the code of the function

Just type the function:

Example with the function ddply from plyr

library(plyr)
> ddply
function (.data, .variables, .fun = NULL, ..., .progress = "none", 
    .inform = FALSE, .drop = TRUE, .parallel = FALSE, .paropts = NULL) 
{
    if (empty(.data)) 
        return(.data)
    .variables <- as.quoted(.variables)
    pieces <- splitter_d(.data, .variables, drop = .drop)
    ldply(.data = pieces, .fun = .fun, ..., .progress = .progress, 
        .inform = .inform, .parallel = .parallel, .paropts = .paropts)
}
<environment: namespace:plyr>

where:

Support

Error: could not find function

To avoid the below message (function)

Error: could not find function "...."

load the package that contains the function

require(myPackage)

with R - Require or R - Library