CODEWITHSUNDEEP
rdataframe
cianius
cianius

Reputation: 2412

take last filled cell in every row in a data.frame R

I have a data frame that has 500 rows and at least a dozen columns per row. Some rows have a lot more columns. What I want to do is to create a variable that contains the contents of the last filled cell, per row.

dput(head(calls_with_overlap, 10))

    A      B        C       D
1. aknasd kasdknsd kasdkas kasdpa
2. asdhad none
3. apihsd piahsdas ashpidapishd phasd askdn wads
4. kasdas none

What I tried so far is: 

if (calls_with_overlap$Overlap[10] = "none") 
{
Ending <- TRUE
} 
else (Ending <- FALSE)
 if (Ending == FALSE) 
   {
  for (q in 1:1024) #max is 1024 columns. perhaps not best way for number of columns.
  {
   EndingCell <- c(10 + q)
   EndingCellcontents <- calls_with_overlap[h,EndingCell] # h is from an outer loop, going through each row. 
     if (EndingCellcontents != ""){ 
       EndingCellcontents[q] <-    EndingCellcontents}
       last  <-  length(EndingCellcontents)
       last<-  EndingCellcontents[last]
      }}

I would like "last" to contain the last filled cell in each row, unless it says "none". Starting from column B.

i.e. in this case

last
[1] kasdpa
[2] wads

Apologies if this is a nooby question, but I'm teaching myself programming from scratch.

Upvotes: 3

Views: 317

Answers (1)

Julius Vainora
Julius Vainora

Reputation: 48211

Edited: mistaken rows and columns

df <- data.frame(A=c("r","","z"),B=c("w","",""))
df
  A B
1 r w
2    
3 z  
v <- apply(df, 1, function(x) x[rev(which((x!="")==TRUE))[1]])
[1] "w" NA  "z"
v[!is.na(v)]
[1] "w" "z"

Upvotes: 4

Related Questions