Passing column names to custom function

Question

I'm trying to make a custom function to correct some answer given the format, number set, and problem of the question. For that I have to identify those three variables from three different columns and then check the answer for the correspondent row.

Consider the next data frame as example of the one I'm using.

df = data.frame(ID = 1:10,
            Formato_SCREE1 = c("FN", "FN", "FN", "PR", "PR", "FN", "PR", "PR", "FN", "PR"),
            Problema_SCREE1 = c("UNIDAD", "DECIMA", "UNIDAD", "UNIDAD", "DECIMA", "DECIMA", "UNIDAD", "DECIMA", "DECIMA", "UNIDAD"),
            Set_SCREE1 = c("SET4", "SET1", "SET4", "SET3", "SET3", "SET4", "SET3", "SET2", "SET1", "SET2"),
            Resp_SCREE1 = c(0.7777778,  0.5000000,  0.7777778,  0.7142857,  2.5000000,  0.7777778, 0.7142857, 0.2857143, 0.3333333, 110.1111111))

ID Formato_SCREE1 Problema_SCREE1 Set_SCREE1 Resp_SCREE1
1   1             FN          UNIDAD       SET4   0.7777778
2   2             FN          DECIMA       SET1   0.5000000
3   3             FN          UNIDAD       SET4   0.7777778
4   4             PR          UNIDAD       SET3   0.7142857
5   5             PR          DECIMA       SET3   2.5000000
6   6             FN          DECIMA       SET4   0.7777778
7   7             PR          UNIDAD       SET3   0.7142857
8   8             PR          DECIMA       SET2   0.2857143
9   9             FN          DECIMA       SET1   0.3333333
10 10             PR          UNIDAD       SET2 110.1111111

My first solution to correct the answer was the following.

temp1 = df %>% 
  filter(Set_SCREE1 == "SET1") %>% 
  mutate(Error_SCREE1 = ifelse( Formato_SCREE1 == "PR" & Problema_SCREE1 == "DECIMA", Correct_answer_PR_DECIMA_SET1 - Resp_SCREE1,
                                ifelse( Formato_SCREE1 == "PR" & Problema_SCREE1 == "UNIDAD", Correct_answer_PR_UNIDAD_SET1 - Resp_SCREE1,
                                        ifelse(Formato_SCREE1 == "FN" & Problema_SCREE1 == "DECIMA", Correct_answer_FN_DECIMA_SET1 - Resp_SCREE1,
                                               ifelse(Formato_SCREE1 == "FN" & Problema_SCREE1 == "UNIDAD", Correct_answer_FN_UNIDAD_SET1 - Resp_SCREE1, 0))))) %>% 
  select(ID, Error_SCREE1)

And the correct answers:

Correct_answer_PR_DECIMA_SET1 = 1
Correct_answer_PR_UNIDAD_SET1 = 2
Correct_answer_FN_DECIMA_SET1 = 3
Correct_answer_FN_UNIDAD_SET1 = 4

This works fine but I have to repeat it four times (for each Set_SCREE1), and then repeat that four times more (for ...SCREE2, ...SCREE3, and ...SCREE4). That result in 16 chunks of code. Too many lines and something difficult to read.

Then I tried to make a function that could do the same in less lines, and end up with this:

error_calculator = function(data, set, formato, problema, correcta) {

temp = data %>% 
filter(Set_SCREE1 == set) %>% 
mutate(Error_SCREE1 = ifelse( formato == "PR" & problema == "DECIMA", correcta - Resp_SCREE1, 
                          ifelse( formato == "PR" & problema == "UNIDAD", correcta - Resp_SCREE1, 
                                      ifelse( formato == "FN" & problema == "DECIMA", correcta - Resp_SCREE1, 
                                              ifelse( formato == "FN" & problema == "UNIDAD", correcta - Resp_SCREE1, 0))))) %>%

select(ID, Error_SCREE1)

return(temp)
}

temp1 = error_calculator(df, "SET1", "Formato_SCREE1", "Problema_SCREE1", "Correct_answer_PR_DECIMA_SET1")  

The problem with this function is that is only working until the filter() line, and then no condition of the following ifelses is accomplished, so the Error_SCREE1 column is filled with "0". My first thought was that you can't pass column names to a function in the way I'm doing it. But according to this you actually can. So I can't really tell why the column names are not being recognized.