CODEWITHSUNDEEP
r
Ariya Woodard
Ariya Woodard

Reputation: 59

create a variable from multiple variables in R

I have a data set looks like, dat. I want to create one new variable (newVar) with values 1 through 4 using the Q1 through Q6 in R. Can anyone help how to do this, please?

newVar of 1: If any variable in Q1 through Q6 = "A"
newVar of 2: **If none of variables in Q1 through Q6 = "A", and** If any variable in Q59_1 through Q6= "B"
newVar of 3: **If none of variables in Q1 through Q6 = "A" or "B", and** If any variable in Q1 through Q6= "C"
newVar of 4: If none of variables in Q1 through Q6 = "A" or "B" or "C" , and**strong text** If any variable in Q1 through Q6= "D"  

dat = data.frame(Q1 = c("A","D","B","A","D","C","B"),
                 Q2 = c("C","A","B","A","C","C","B"),
                 Q3 = c("B","D","D","A","B","C","A"),
                 Q4 = c("A","C","A","D","B","D","B"),
                 Q5 = c("C","D","D","A","A","C","B"),
                 Q6 = c("D","B","A","A","C","C","A"),
                 Q7 = c("5","1","2","2","4","0","3"),
                 Q8 = c("T","U","O","O","Z","W","Y"),
                 Q9 = c("X","T","V","P","I","U","Y"))

Additional comment: I actually need all the conditions especially the bold parts (If none of variables in ...) such as NewVar 2: If none of variables in Q1 through Q6 = "A", and If any variable in Q59_1 through Q6= "B" 3: If none of variables in Q1 through Q6 = "A" or "B", and If any variable in Q1 through Q6= "C" 4: If none of variables in Q1 through Q6 = "A" or "B" or "C" , and If any variable in Q1 through Q6= "D"

Upvotes: 0

Views: 674

Answers (3)

GKi
GKi

Reputation: 39647

You can use match with LETTERS and min to test if any of the LETTERS is present.

dat$newVar <- apply(dat[1:6], 1, function(x) min(match(x, LETTERS[1:4])))
dat
#  Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 newVar
#1  A  C  B  A  C  D  5  T  X      1
#2  D  A  D  C  D  B  1  U  T      1
#3  B  B  D  A  D  A  2  O  V      1
#4  A  A  A  D  A  A  2  O  P      1
#5  D  C  B  B  A  C  4  Z  I      1
#6  C  C  C  D  C  C  0  W  U      3
#7  B  B  A  B  B  A  3  Y  Y      1

Upvotes: 1

Ronak Shah
Ronak Shah

Reputation: 388817

In base R, you can use apply row-wise on selected columns and assign value to newVar checking conditions one by one. 

cols <- paste0('Q', 1:6)

dat$newVar <- apply(dat[cols], 1, function(x) {
    if(any(x == 'A')) 1
    else if(any(x == 'B')) 2
    else if(any(x == 'C')) 3
    else if(any(x == 'D')) 4
    else NA
})

dat
#  Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 newVar
#1  A  C  B  A  C  D  5  T  X      1
#2  D  A  D  C  D  B  1  U  T      1
#3  B  B  D  A  D  A  2  O  V      1
#4  A  A  A  D  A  A  2  O  P      1
#5  D  C  B  B  A  C  4  Z  I      1
#6  C  C  C  D  C  C  0  W  U      3
#7  B  B  A  B  B  A  3  Y  Y      1

Upvotes: 0

Martin Gal
Martin Gal

Reputation: 16978

Using packages tidyr and dplyr:

dat %>%
  mutate(id = row_number()) %>%
  pivot_longer(cols=starts_with("Q"), names_to="Q", names_prefix="Q") %>%
  group_by(id) %>%
  mutate(newVar=case_when(Q %in% 1:6 & any(value == "A") ~ 1,
                          Q %in% 1:6 & any(value == "B") ~ 2,
                          Q %in% 1:6 & any(value == "C") ~ 3,
                          Q %in% 1:6 & any(value == "D") ~ 4)) %>%
  pivot_wider(names_from="Q", names_prefix="Q", names_sep="") %>%
  fill(everything(), .direction="down") %>%
  slice(2)

I'm not sure if I got your conditions right. And there are surely more tidy ways, but this one yields

# A tibble: 7 x 11
# Groups:   id [7]
     id newVar Q1    Q2    Q3    Q4    Q5    Q6    Q7    Q8    Q9   
  <int>  <dbl> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
1     1      1 A     C     B     A     C     D     5     T     X    
2     2      1 D     A     D     C     D     B     1     U     T    
3     3      1 B     B     D     A     D     A     2     O     V    
4     4      1 A     A     A     D     A     A     2     O     P    
5     5      1 D     C     B     B     A     C     4     Z     I    
6     6      3 C     C     C     D     C     C     0     W     U    
7     7      1 B     B     A     B     B     A     3     Y     Y

Upvotes: 1

Related Questions