Rename factor levels based on condition in R

Question

Let's say that we have the following dataframe:

data1 <- data.frame(S1 = sample(c("A", "B", "C"), size = 20, replace = TRUE),
                    S2 = sample(c("A", "B", "C"), size = 20, replace = TRUE),
                    number = rnorm(20, 10, 100))

There's also this variable:

scenario <- "scenario1" # can take one of thre values: 'scenario1', 'scenario2', 'scenario3'

There are also the following mappings. Please note that level names in different scenarios can't be handled automatically, eg with a regex:

changer_1 <- c("A"="a1", "B"="b1", "C"="c1")
changer_2 <- c("A"="something", "B"="completely", "c"="different")
changer_3 <- c("A"="z2", "B"="d3", "c"="p14")

I want to rename the factor levels based on the scenario 's value (condition). What I've come up so far is a rather primitive if/else block that checks the condition and then uses one of the changer s:

if(scenario=="scenario1"){
  data1$S1 <- revalue(data1$S1, changer_1)
  data1$S2 <- revalue(data1$S2, changer_1)
} else if (scenario=='scenario2'){
  data1$S1 <- revalue(data1$S1, changer_2)
  data1$S2 <- revalue(data1$S2, changer_2)
} else {
  data1$S1 <- revalue(data1$S1, changer_3)
  data1$S2 <- revalue(data1$S2, changer_3)
}

I wonder whether there is other (more R-like) way to do this?

Answer 1

One way to do this is to build a list with all the "scenario"'s and their mappings

change_list <- list(scenario1 =  c("A"="a1", "B"="b1", "C"="c1"), 
             scenario2 = c("A"="something", "B"="completely", "C"="different"), 
             scenario3 = c("A"="z2", "B"="d3", "C"="p14"))

Now create a function which returns the values as per scenario

get_values <- function(change_list, scenario, x) {
   change_list[[scenario]][x]
}

and now you can call the function

get_values(change_list, "scenario1", data1$S1)

#  A    C    B    B    B    C    B    B    C    A    A    C    C    A    B  ...  
#"a1" "c1" "b1" "b1" "b1" "c1" "b1" "b1" "c1" "a1" "a1" "c1" "c1" "a1" "b1" ...

get_values(change_list, "scenario2", data1$S1)

#          A            C            B            B            B            C ..   
#  "something"  "different" "completely" "completely" "completely"  "different"..

In this way with only one function you can change all the values in any column.