Reputation: 139
I'm trying to analyze the efficiency of the lapply function in R, but I didn't find the original developing code of lapply in the folder "base", so I wrote a simple code by myself.
new_lapply<-function(data,f_name){
list<-list()
for (i in length(data)){
list[i]<-f_name(data[,i])
}
return (list)
}
check<-Boston[1:10,]
lapply(check,length)
new_lapply(check,length)
Error: could not find function "f_name"
I wonder how to input a "general function name" in to a function, so I could run different functions in the new_lapply function, similar as the built-in lapply().
Thank you so much.
Upvotes: 2
Views: 95
Reputation: 99331
You should use match.fun()
. I've made a couple of other changes too. Hopefully this gets you in the right direction. Also, there's a good explanation of lapply()
and its efficiency here
new_lapply <- function(data, f_name) {
## match the function in the 'f_name' argument
f_name <- match.fun(f_name)
## allocate a list the same length as 'data' (ncol for data frames)
List <- vector("list", length(data))
for (i in seq_along(data)) {
List[[i]] <- f_name(data[[i]])
}
## if 'data' has names, carry them over to 'List'
if(!is.null(names(data)))
names(List) <- names(data)
List
}
identical(lapply(mtcars, length), new_lapply(mtcars, length))
# [1] TRUE
um <- unname(mtcars)
identical(lapply(um, length), new_lapply(um, length))
# [1] TRUE
Upvotes: 1