Recode a dataframe variable based on matches in a list

Question

I am trying to recode a variable in a data frame based on matches to elements in a separate list. For example:

df <- data.frame(stringsAsFactors = FALSE,
  var1 = c("116", "117", "118", "SL1", "SL2", "234"))

matchList <- list(c("116, 117, and 118", "116", "117", "118"), 
c("SL1/SL2", "SL1", "SL2"))

df
var1
1     116
2     117
3     118
4     SL1
5     SL2
6     234

matchList
[[1]]
[1] "116, 117, and 118" "116"               "117"               "118"              

[[2]]
[1] "SL1/SL2" "SL1"     "SL2"

If the original var1 element matches items 2 - 4 of a matchList element, it should be recoded with item 1 of that same list element. I want the recoded var1 to look like the following:

var1
1     116, 117, and 118
2     116, 117, and 118
3     116, 117, and 118
4     SL1/SL2
5     SL1/SL2
6     234

The following lines of code work one list element at a time, but I'm not clear on how to automate this:

# get indices of matches for matchList element 1
r <- which(df$var1 %in% matchList[[1]]) 
# replace matches with first list item of list element 1 using indices of matches
df$var1[r] <- matchList[[1]][1]

I've tried the following for loop, but I'm not sure what I'm missing

for (i in length(matchList)){
  r <- which(df$var1 %in% matchList[[i]])
  df$var1[r] <- matchList[[i]][1]
}

score 0 · Accepted Answer · answered Jul 17 '18 at 03:07

The issue is in the length(matchList) which is a single value i.e. 2. Instead, we need to loop through the sequence

for(i in seq_along(matchList)) {
     r <- which(df$var1 %in% matchList[[i]])
     df$var1[r] <- matchList[[i]][1]
  } 
df
#               var1
#1 116, 117, and 118
#2 116, 117, and 118
#3 116, 117, and 118
#4           SL1/SL2
#5           SL1/SL2
#6               234

score 0 · Answer 2 · answered Jul 17 '18 at 03:22

An alternative to your approach is to do the same thing with lapply using the <<- operator.

lapply(matchList, function(x) {
  df$var1[df$var1 %in% x] <<- x[[1]]
})
df

#              var1
#1 116, 117, and 118
#2 116, 117, and 118
#3 116, 117, and 118
#4           SL1/SL2
#5           SL1/SL2
#6               234

However, before using <<- you might want to read How do you use scoping assigment.

Recode a dataframe variable based on matches in a list

2 Answers2