How do add a column in a data frame in R

Question

I have imported data from a file into a data frame in R. It is something like this.

Name      Count   Category
A         100     Cat1
C         10      Cat2
         


        

Answer 1

check out:

• cut()
• recode() in the car package

Answer 2

Perhaps simpler and more readable using ifelse and %in%:

df <- data.frame( Name = c('A', 'C', 'D', 'E', 'H', 'Z', 'M'), 
Count =c(100,10,40,30,3,20,50), stringsAsFactors = FALSE)

cat1 = c("A","D")
cat2 = c("C","Z")
cat3 = c("E","H")
cat10 = c("M")

df$Category = ifelse(df$Name %in% cat1, "Cat1",
              ifelse(df$Name %in% cat2, "Cat2",
              ifelse(df$Name %in% cat3, "Cat3",
              ifelse(df$Name %in% cat10, "Cat10",
              NA))))

   Name Count Category
1    A   100     Cat1
2    C    10     Cat2
3    D    40     Cat1
4    E    30     Cat3
5    H     3     Cat3
6    Z    20     Cat2
7    M    50    Cat10

Answer 3

[Update following the OP's comment and altered Q]

DF <- data.frame(Name = c("A","C","D","E","H","Z","M"),
                 Count = c(100,10,40,30,3,20,50), stringsAsFactors = FALSE)
lookup <- data.frame(Name = c("A","C","D","E","H","Z","M"),
                     Category = paste("Cat", c(1,2,1,3,3,2,10), sep = ""),
                     stringsAsFactors = FALSE)

Using the above data frames, we can do a data base merge. You need to set-up lookup for the Name Category combinations you want, which is OK if there aren't a very large number of Names (At least you only need to list them once each in lookup and you don't have to do it in order - list all Cat1 Names first, etc):

> merge(DF, lookup, by = "Name")
  Name Count Category
1    A   100     Cat1
2    C    10     Cat2
3    D    40     Cat1
4    E    30     Cat3
5    H     3     Cat3
6    M    50    Cat10
7    Z    20     Cat2
> merge(DF, lookup, by = "Name", sort = FALSE)
  Name Count Category
1    A   100     Cat1
2    C    10     Cat2
3    D    40     Cat1
4    E    30     Cat3
5    H     3     Cat3
6    Z    20     Cat2
7    M    50    Cat10

---

One option is indexing:

foo <- function(x) {
    out <- character(length = length(x))
    chars <- c("Ones", "Tens", "Hundreds", "Thousands")
    out[x < 10] <- chars[1]
    out[x >= 10 & x < 100] <- chars[2]
    out[x >= 100 & x < 1000] <- chars[3]
    out[x >= 1000 & x < 10000] <- chars[4]
    return(factor(out, levels = chars))
}

An alternative that scales better is,

bar <- function(x, cats = c("Ones", "Tens", "Hundreds", "Thousands")) {
    out <- cats[floor(log10(x)) + 1]
    factor(out, levels = cats)
}

Answer 4

You can use ifelse. If your data frame were called df you would do:

df$cat <- ifelse(df$name<100, "Ones", "Hundreds")
df$cat <- ifelse(df$name<1000, df$cat, "Thousands")

Answer 5

You can use a map. (UPDATED to use stringsAsFactors = FALSE)

df <- data.frame( Name = c('A', 'C', 'D', 'E', 'H', 'Z', 'M'), 
                  Count = c(100,10,40,30,3,20,50), stringsAsFactors = FALSE)
Categories <- list(Cat1 = c('A','D'), 
                   Cat2 = c('C','Z'), 
                   Cat3 = c('E','H'), 
                   Cat10 = 'M')
nams <- names( Categories )
nums <- sapply(Categories, length)
CatMap <- unlist( Map( rep, nams, nums ) )
names(CatMap) <- unlist( Categories )

df <- transform( df, Category = CatMap[ Name ])