Change Responses to 1 and NA to 2-CodePudding

How do you rename any response/value to 1 and any NA to 0?

Sample Dataframe:

Dataframeexample <- data.frame(Q5 = c("potato", "chips", "chips", "chips,potato","icecream,chips,potato", "icecream,potato", "chips", "NA", "NA"))

My actual dataframe has hundreds of potential combinations, so renaming each potential value individually isn't feasible.

CodePudding user response：

Do you mean like this? I fixed your NA values to be logical rather than characters.

#### Data Fix ####
library(tidyverse) # for mutate later

Dataframeexample <- data.frame(Q5 = c("potato", "chips", "chips",
                                      "chips,potato","icecream,chips,potato",
                                      "icecream,potato",
                                      "chips", NA, NA))
#### Ifelse Statement ####
Dataframeexample %>% 
  mutate(Q5 = ifelse(is.na(Q5),
                     0,
                     1))

Giving you this data:

CodePudding user response：

First of all, if you really have that data, convert "NA" strings to real NA.

dat[dat == "NA"] <- NA

Then rename values in the entire data frame at once:

dat[!is.na(dat)] <- 1
dat[is.na(dat)] <- 0
dat
#   Q5 Q6
# 1  1  1
# 2  1  1
# 3  1  1
# 4  0  0
# 5  0  0

No need for packages or loops.

Data:

dat <- data.frame(Q5 = c("potato", "chips", "chips,potato", "NA", "NA"),
                  Q6 = c("potato", "chips", "chips,potato", NA, NA))

CodePudding user response：

Write a function to do the conversions and apply it to the data set's character columns. The function below works even for the true NA's present in the data.

na_falsetrue <- function(x, value = "NA") {
  is.na(x) <- x == value
  as.integer(!is.na(x))
}

i_char <- sapply(Dataframeexample, is.character)
Dataframeexample[i_char] <- lapply(Dataframeexample[i_char], na_falsetrue)

Dataframeexample
#>   Q5
#> 1  1
#> 2  1
#> 3  1
#> 4  1
#> 5  1
#> 6  1
#> 7  1
#> 8  0
#> 9  0

^{Created on 2022-10-30 with reprex v2.0.2}