R在数据帧内将字符串转换为整数

Question

REF         ALT 
 AT  ATT,A,ATTT

I'm working with the data frame above and need to convert the REF column containing just AT into an integer of 0 and then the ALT column containing ATT,A,ATTT into 1,2,3 , I have tried Transform but it didn't seem to work.

Also my full data frame has multiple rows so I would need to loop the command to apply to all rows.

Any help would be greatly appreciated

Answer 1

Loop in R is really inefficient so I would strongly advise against it if it is not absolutely necessary and in your case I don't think that it is necessary.

for example you can do, (assuming your dataframe is called "df"):

df$REF<-0 # from what I gather all rows should be 0


df$ALT2<-1 # a proxy column that you can copy to ALT after
df$ALT2[df$ALT == "A"] <-2 # converts A to 2
df$ALT2[df$ALT == "ATTT"] <-3 # converts ATTT to 2

df$ALT<-df$ALT2 # copy proxy over to ALT
df$ALT2<-NULL #erase proxy column

If you don't care which character string gets assigned to which number, but you just want different strings to have a different intetger, you can also do:

df$REF<-0 # from what I gather all rows should be 0
df$ALT <- as.numeric(factor(df$ALT)) # give a distinct number to each distinct string counting up from 1.

Answer 2

Setting the REF column to 0 is straightforward.

df$REF <- 0

For the ALT column I assume that order for each entry matters but rows are independent. So an A could be numbered 1 in one row but 2 in another (if there are multiple entries in that row). So the only thing we care about is the number of alternatives in each row. We can simply count them and generate a vector with the appropriate numbers, collapsing them into a single string to form the corresponding entry in the data frame:

df$ALT <- sapply(strsplit(df$ALT, ","), 
    function(alt) paste(1:length(alt), collapse=","))