How to subset a row from list based on condition - r

Question

I have a data.table outcome, which has a column called hospital. and a column called state. The outcome has been sorted already. Now I want to subset the nth hospital from each state(if there's not a Nth then returns NA for that state). I try to solve it the below way.(Since this is a homework, I just showed the third branch that causes mistake).
rankall <- function(out, num = "best"){
outcome <- readdata(outcome = out) //returns a data.table sorted by rate
...
outcome <- lapply(outcome, function(x) ifelse(num <= nrow(x), x[num,], c(NA,NA)))
outcome <- rbindlist(outcome)
}
The original outcome is like
> data
hospital state
1: NYU HOSPITALS CENTER NY
2: DOYLESTOWN HOSPITAL PA
3: AVERA HEART HOSPITAL OF SOUTH DAKOTA LLC SD
4: GLENDALE ADVENTIST MEDICAL CENTER CA
5: WATERBURY HOSPITAL CT
---
2716: DESERT SPRINGS HOSPITAL NV
2717: THREE RIVERS COMMUNITY HOSPITAL OR
2718: ROBERT WOOD JOHNSON UNIVERSITY HOSPITAL AT RAHWAY NJ
2719: LAREDO MEDICAL CENTER TX
2720: MEDICAL CENTER SOUTH ARKANSAS AR
And the first and second branch could produce the right result, which is like
> head (data)
hospital state
1: NA AK
2: CRESTWOOD MEDICAL CENTER AL
3: ARKANSAS HEART HOSPITAL AR
4: MAYO CLINIC HOSPITAL AZ
5: GLENDALE ADVENTIST MEDICAL CENTER CA
6: ST MARYS HOSPITAL AND MEDICAL CENTER CO
> nrow(data)
[1] 54
However, the third condition just COULDN'T do its work. Which produce the error
Error in rbindlist(outcome) :
Item 1 of list input is not a data.frame, data.table or list
And after debugging I found out that the outcome after the condition is something like(which caused the error in the last step)
$AK
[1] NA
$AL
$AL[[1]]
[1] "HIGHLANDS MEDICAL CENTER"
Differs from the first two which is like...
> head(data,2)
$AK
hospital state
1: PROVIDENCE ALASKA MEDICAL CENTER AK
$AL
hospital state
1: CRESTWOOD MEDICAL CENTER AL
So I wonder what's wrong with the third branch.
Could anyone help me out, thank you very much!!!
By the way, I wonder if I could refer to a variable with the same name of another's. Such like when I called readdata, I need to pass a argument called outcome, which prevents me to use this name as the argument of the rankall function(I use out instead). I know in JAVA this.outcome will help, so how about in R.

Answer 1

Thank you for Vivek's help, I have figured it out now.
First is about the mis-performance of the third branch. It'll work properly if I first convert numto a number using as.numeric(num). I think it's because the num is regarded as a character(since it has possible value of "best" and "worst") that caused the mistake.
The second regarding the naming space is however strange. Though not having performed well in my own test, it just do work after Vivek answer my question. So, that means we can just use the following code, and R will get us the right result.
rankall <- function(outcome, num = "best"){
outcome <- readdata(outcome = outcome)

Answer 2

This output below is a warning sign, first element of outcome is just NA, an atomic which is forced to 'bind' with second element which is a list $AL[[1]] by the function rbindlist
$AK
[1] NA
$AL
$AL[[1]]
[1] "HIGHLANDS MEDICAL CENTER"
Possible solution:
One way would be to have data.frame or list output in all three cases if condition blocks e.g. with data.frame, once num is numeric the following should work in all cases, best,worst and intermediate row.
sapply(outcome, function(x) data.frame(state=x[num,"state"],hospital=x[num,"hospital"])
for cases with no matches the hospital column should have an NA, could you please check if this works