remove all variables except functions

Question

I have loaded in a R console different type of objects. I can remove them all using

rm(list=ls())

or remove only the functions (but not the variables) using

rm(list=lsf.str())

My question is: is there a way to remove all variables except the functions

Answer 1

You can use the following command to clear out ALL variables. Be careful because it you cannot get your variables back.

rm(list=ls(all=TRUE))

Answer 2

Here's a one-liner that removes all objects except for functions:

rm(list = setdiff(ls(), lsf.str()))

It uses setdiff to find the subset of objects in the global environment (as returned by ls()) that don't have mode function (as returned by lsf.str())

Answer 3

Here's a pretty convenient function I picked up somewhere and adjusted a little. Might be nice to keep in the directory.

list.objects <- function(env = .GlobalEnv) 
{
    if(!is.environment(env)){
        env <- deparse(substitute(env))
        stop(sprintf('"%s" must be an environment', env))
    }
    obj.type <- function(x) class(get(x, envir = env))
    foo <- sapply(ls(envir = env), obj.type)
    object.name <- names(foo)
    names(foo) <- seq(length(foo))
    dd <- data.frame(CLASS = foo, OBJECT = object.name, 
                     stringsAsFactors = FALSE)
    dd[order(dd$CLASS),]
}

> x <- 1:5
> d <- data.frame(x)
> list.objects()
#        CLASS       OBJECT
# 1 data.frame            d
# 2   function list.objects
# 3    integer            x 
> list.objects(env = x)
# Error in list.objects(env = x) : "x" must be an environment

Answer 4

The posted setdiff answer is nice. I just thought I'd post this related function I wrote a while back. Its usefulness is up to the reader :-).

lstype<-function(type='closure'){ 
    inlist<-ls(.GlobalEnv)
    if (type=='function') type <-'closure'
    typelist<-sapply(sapply(inlist,get),typeof)
    return(names(typelist[typelist==type]))
}