Using paste in data frame column call

Question

combine_cols<- function(primary,secondary,linker,column) {
require(data.table) 
a<-data.table("Sample"=primary[,linker], primary[,column])
b<-data.table("Sample"=secondary[,linker], secondary[,column])

c <- merge(a, b, by = "Sample", all=TRUE)
c[,Status := ifelse(!is.na(c[,paste0(column,".x")]), paste0(column,".x"), 
paste0(column,".y"))]
c[,`:=` (paste0(column,".x")=NULL, paste0(column,".y")= NULL)]

return(c)
}
mydata1<-data.frame("Sample"=c("100","101","102","103"),"Status"=c("Y","","","partial"))
mydata2<-data.frame("Sample"=c("100","101","102","103","106"),"Status"=c("NA","Y","","","Y"))
print((combine_cols(mydata1,mydata2,"Sample",c("Status"))))

I'm trying to create a function to merge columns of split data. The ifelse line isn't working because the paste0(column,".x") is recognized as a character and not a column name. How can I ensure that c[,paste0(column,".x")] reflects c$c[,paste0(column,".x")] ? Better yet, how can I modify this line to handle a list of column names?

when naming an object, please avoid R functions, like c, which is used in your code. — Sean Lin, Jun 14 '17 at 15:20
Possible dupe: [Using dynamic column names with `data.table`](https://stackoverflow.com/q/14937165/903061) — Gregor Thomas, Jun 14 '17 at 15:30

score 0 · Answer 1 · answered Jun 14 '17 at 16:42

Just work with standard names and rename after, it'll be much more readable too.

a<-data.table("Sample"=primary[,linker], "tempname" =primary[,column])        # added tempname
b<-data.table("Sample"=secondary[,linker], "tempname" =secondary[,column])    # added tempname
c <- merge(a, b, by = "Sample", all=TRUE)
c[,Status := ifelse(!is.na(tempname.x),tempname.x,tempname.y)]
setnames(c,paste0("tempname",c(".x",".y")),paste0(column,c(".x",".y")))

with your example:

   Sample Status.x Status.y Status
1:    100        Y       NA      3
2:    101                 Y      1
3:    102                        1
4:    103  partial               2
5:    106       NA        Y      3

I have no idea what the following line (before return) is supposed to do, and it will fail, but as it's not part of the question (yet), here goes.

Using paste in data frame column call

1 Answers1