Question

This might be very trivial but I could not find an easy solution anywhere. I am trying to create a script in R to count entries in one column that belong to a one of 3 categories specified another column. I have a list of clinical patients with ID numbers (more than one entry for the same ID) that have been seen by two services (a or b). I need to know how many ID have been seen by service a and by service b and service c, but counting repeated visits by a service only once (so basically the number of patients that have used each service at least once) – hope this makes sense, here is an example to explain.

Example:

     ID    Category

     A001  a

     A002  a

     A002  a

     A002  b

     A003  b

     A003  b

     A005  c

     A001  a

     A004  b

     A004  b

     A006  c

     A006  a

Output should be something like:
     a=3
     b=3
     c=2

This is what I have done, but I am quite stuck, and this might not be good at all!
 DataString<- matrix(nrow=dim(refnum)[1], ncol=1)
 for (i in 1:dim(refnum)[1]){
   DataString[i,1]<- paste(refnum[i,], collapse = '')
 }

 #generate vector of unique strings
 uniqueID<- unique(DataString)

 #create new matrix to store new IDs
 newID<- matrix(nrow=dim(data)[1], ncol=1)

 #initiate index n
 n<-0
 #loop through unique strings
 for (i in 1:dim(refnum)[1]){
   #increment n by 1 for each increment through unique strings
   n<- n+1
   #loop through data rows
   for (j in 1:dim(data)[1]){    
     #find matches with string i
     index<- which(DataString == uniqueID[i,1])
     #assign new ID to matching rows
     newID[index,1]<- n
   }
 }

												

Answer 1

One of the many solutions:

table(df[!duplicated(df), "Category"])

# a b c 
# 3 3 2