Is there a way I can make a dummy variable for multipule varaibles? Trying to have "White"-CodePudding

I tried this;

new_data $Ethnicity_dummy = as.numeric(new_data$Ethnicity == "White"|"White British"| "White Other")

but I get this error message;

new_data $Ethnicity_dummy = as.numeric(new_data$Ethnicity == "White"|"White British"| "White Other")
Error in new_data$Ethnicity == "White" | "White British" : 
  operations are possible only for numeric, logical or complex types

CodePudding user response：

You are using the "or" (|) function incorrectly - if you wanted to use the or notation, you would have to specify the variable for each time (i.e. new_data$Ethnicity == "White"|new_data$Ethnicity =="White British"...)

A much easier way is %in% - try using:

new_data$Ethnicity_dummy = as.numeric(new_data$Ethnicity %in% c("White","White British","White Other"))

CodePudding user response：

library(tidyverse)

new_data <- tibble(
  ethnicity = sample(c("white", "white british", "white other"), 
                     10, replace = TRUE)
)

# A tibble: 10 x 1
   ethnicity    
   <chr>        
 1 white        
 2 white        
 3 white british
 4 white other  
 5 white british
 6 white british
 7 white british
 8 white british
 9 white        
10 white        

new_data %>%  
  mutate(dummy = ethnicity %>% 
           as_factor() %>% 
           as.numeric())

# A tibble: 10 x 2
   ethnicity     dummy
   <chr>         <dbl>
 1 white             1
 2 white             1
 3 white british     2
 4 white other       3
 5 white british     2
 6 white british     2
 7 white british     2
 8 white british     2
 9 white             1
10 white             1