Replace NA in a series of variables with different types of missing

问题

This is my data.

# A tibble: 10 x 6
     id  main   s_0   s_1   s_2   s_3  
   <dbl> <fct> <fct> <fct> <fct> <fct>
 1    1     5    75    A     4     110  
 2    2    NA    NA    NA    NA    NA  
 3    3    11    13    NA    7     769  
 4    4    NA    NA    NA    NA    NA   
 5    5    11    NA    NA    NA    835  
 6    6    13    39    NA    4     NA   
 7    7    NA    NA    NA    NA    NA  
 8    8    19    42    D     6     654   
 9    9    20    4     NA    7     577  
10   10    NA    NA    NA    NA    NA

As you can see, the column main indicates that rows in the other columns (s_0: s_4) answered the questions or not. Ids 2,4,7 and 10 were not eligible for the rest, however, other participants can answer or miss (s_0:s_4). So I have a mix of NAs, I want to use a code that can identify the source of missing. The code that I am using mix all kind of missing :

library(dplyr)
library(forcats)

# Make sample data vars factors
dat <- dat %>%
  mutate(across(starts_with("s_"), as.factor))

# Add 'No' as factor level
dat %>%
  mutate(across(starts_with("s_"), fct_explicit_na, "No"))

While I want to have something like this:

# A tibble: 10 x 6
     id  main   s_0   s_1   s_2   s_3  
   <dbl> <fct> <fct> <fct> <fct> <fct>
 1    1     5    75    A     4     110  
 2    2    NO1   NO1   NO1   NO1   NO1  
 3    3    11    13    NO    7     769  
 4    4    NO1   NO1   NO1   NO1   NO1   
 5    5    11    NO    NO    NO    835  
 6    6    13    39    NO    4     NA   
 7    7    NO1   NO1   NO1   NO1   NO1 
 8    8    19    42    D     6     654   
 9    9    20    4     NO    7     577  
10   10    NO1   NO1   NO1   NO1   NO1

回答1:

Try :

#Convert columns to characters
df[-1] <- lapply(df[-1], as.character)
#Find index of `NA` value in `main` column
inds <- is.na(df$main)
#Change all the columns to "NO1" in row inds
df[inds, -1] <- 'NO1'
#Change remaining NA values to "NO"
df[is.na(df)] <- 'NO'
df

#   id main s_0 s_1 s_2 s_3
#1   1    5  75   A   4 110
#2   2  NO1 NO1 NO1 NO1 NO1
#3   3   11  13  NO   7 769
#4   4  NO1 NO1 NO1 NO1 NO1
#5   5   11  NO  NO  NO 835
#6   6   13  39  NO   4  NO
#7   7  NO1 NO1 NO1 NO1 NO1
#8   8   19  42   D   6 654
#9   9   20   4  NO   7 577
#10 10  NO1 NO1 NO1 NO1 NO1

来源：https://stackoverflow.com/questions/65095558/replace-na-in-a-series-of-variables-with-different-types-of-missing

标签

database

dataframe

dplyr

missing-data