Question

This is a followup question to this question, initially inspired by this question, but not quite the same.

This is my situation. First I pull some data from a database,

df <- data.frame(id = c(1:6),
                 profession = c(1, 5, 4, NA, 0, 5))
   df
#  id profession
#  1          1
#  2          5
#  3          4
#  4         NA
#  5          0
#  6          5

Second, I pull a key-table with human readable information about the profession codes,

profession.codes <- data.frame(profession.code = c(1,2,3,4,5),
                               profession.label = c('Optometrists',
                               'Accountants', 'Veterinarians', 
                               'Financial analysts',  'Nurses'))                 
   profession.codes
#  profession.code   profession.label
#               1       Optometrists
#               2        Accountants
#               3      Veterinarians
#               4 Financial analysts
#               5             Nurses

Now, I would like to overwrite the profession variable in my df with the labels from profession.codes, preferably using join from the plyr package, but I’m open to any smart solution. Though I do like that ply preserves the order of x.

I currently do it like this,

# install.packages('plyr', dependencies = TRUE)
library(plyr)

profession.codes$profession <- profession.codes$profession.code
df <- join(df, profession.codes, by="profession")
# levels(df$profession.label)
df$profession.label <- factor(df$profession.label, 
   levels = c(levels(df$profession.label), 
   setdiff(df$profession, df$profession.code)))
# levels(df$profession.label)
df$profession.label[df$profession==0 ] <- 0
df$profession.code <- NULL
df$profession  <- NULL
names(df) <- c("id", "profession")
df
#  id         profession
#  1       Optometrists
#  2             Nurses
#  3 Financial analysts
#  4               <NA>
#  5                  0
#  6             Nurses

This is how I overwrite profession without losing the NA and the 0.

The problem is that the 0 could be a 17 or any number and I would like to account for that in some way. Furthermore, I would also like to shorten my code, if possible.

Any help would be greatly appreciated.

Thanks,
Eric

Answer 1

This is one approach in base:

df <- data.frame(id = c(1:6),
                 profession = c(1, 5, 4, NA, 0, 5))

pc <- data.frame(profession.code = c(1,2,3,4,5),
                               profession.label = c('Optometrists',
                               'Accountants', 'Veterinarians', 
                               'Financial analysts',  'Nurses'))  


df$new <- as.character(pc[match(df$profession,  
    pc$profession.code), 'profession.label'])
df[is.na(df$new), 'new'] <- df[is.na(df$new), 'profession'] 
df$new <- as.factor(df$new)
df

Which yields:

  id profession                new
1  1          1       Optometrists
2  2          5             Nurses
3  3          4 Financial analysts
4  4         NA               <NA>
5  5          0                  0
6  6          5             Nurses