CODEWITHSUNDEEP
rdebuggingif-statement
Charlotte
Charlotte

Reputation: 49

Else if to create a new variable DEBUG

I want to create a new variable "influence" in my dataset (f) based on several conditions taking into account other variables. Here is my code:

f$influence <- if (f$ApportLysine_gj <=  f$LysBesoinPlus5 & f$ApportLysine_gj >= f$LysBesoinMoins5){  
f$influence = 1  
} else  if (f$ApportLysine_gj  > f$LysBesoinPlus5 & f$ApportLysine_gj < f$LysExces20){  
 f$influence = 2  
} else  if (f$ApportLysine_gj >= f$LysExces20) {  
f$influence = 3
} else  if (f$ApportLysine_gj < f$LysBesoinMoins5 & f$ApportLysine_gj > f$LysDeficit20){
f$influence = 4
} else {  
f$influence = 5}

I only get an error message "argument is of length zero". What am I doing wrong?

Upvotes: 0

Views: 36

Answers (2)

Rui Barradas
Rui Barradas

Reputation: 76402

You can try using a logical index and assigning the values of the new variable according to its value.

f$influence <- 5
i <- f$ApportLysine_gj <= f$LysBesoinPlus5 & f$ApportLysine_gj >= f$LysBesoinMoins5
f$influence[i] <- 1
i <- f$ApportLysine_gj > f$LysBesoinPlus5 & f$ApportLysine_gj < f$LysExces20
f$influence[i] <- 2
i <- f$ApportLysine_gj >= f$LysExces20
f$influence[i] <- 3
i <- f$ApportLysine_gj < f$LysBesoinMoins5 & f$ApportLysine_gj > f$LysDeficit20
f$influence[i] <- 4

Upvotes: 1

AndS.
AndS.

Reputation: 8110

I often find that if I have lots of if, else if, or ifelse in coding a variable, dplyr's case_when is very handy. Please see below, keep in mind that I couldn't test this without having your data.

library(dplyr)

f %>%
  mutate(influence = case_when(
    ApportLysine_gj <= LysBesoinPlus5 & ApportLysine_gj >= LysBesoinMoins5 ~ 1,
    ApportLysine_gj  > LysBesoinPlus5 & ApportLysine_gj < LysExces20 ~ 2,
    ApportLysine_gj >= LysExces20 ~ 3,
    ApportLysine_gj < LysBesoinMoins5 & ApportLysine_gj > LysDeficit20 ~ 4,
    TRUE ~ 5
  ))

Upvotes: 3

Related Questions