Question

I have a data frame that I’ve discretized using RWeka. RWeka’s discretization creates bins with single quotes in them. Although they are not causing any problems, while plotting it looks ugly to have a variable with 'All' category.

Here’s the discretized data frame:

structure(list(outlook = structure(c(1L, 1L, 2L, 3L, 3L, 3L, 
2L, 1L, 1L, 3L, 1L, 2L, 2L, 3L), .Label = c("sunny", "overcast", 
"rainy"), class = "factor"), temperature = structure(c(1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = "'All'", class = "factor"), 
humidity = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L), .Label = "'All'", class = "factor"), 
windy = c(FALSE, TRUE, FALSE, FALSE, FALSE, TRUE, TRUE, FALSE, 
FALSE, FALSE, TRUE, TRUE, FALSE, TRUE), play = structure(c(2L, 
2L, 1L, 1L, 1L, 2L, 1L, 2L, 1L, 1L, 1L, 1L, 1L, 2L), .Label = c("yes", 
"no"), class = "factor")), .Names = c("outlook", "temperature", 
"humidity", "windy", "play"), row.names = c(NA, -14L), class = "data.frame")

How can I remove the single quotes from the data and recreate the factors?

Answer 1

This should do it:

df$temperature <- gsub("\\'", "", df$temperature)
df$humidity <- gsub("\\'", "", df$humidity)
> df
    outlook temperature humidity windy play
1     sunny         All      All FALSE   no
2     sunny         All      All  TRUE   no
3  overcast         All      All FALSE  yes
4     rainy         All      All FALSE  yes
5     rainy         All      All FALSE  yes
6     rainy         All      All  TRUE   no
7  overcast         All      All  TRUE  yes
8     sunny         All      All FALSE   no
9     sunny         All      All FALSE  yes
10    rainy         All      All FALSE  yes
11    sunny         All      All  TRUE  yes
12 overcast         All      All  TRUE  yes
13 overcast         All      All FALSE  yes
14    rainy         All      All  TRUE   no

If you need to do the same over several columns, this might be more efficient.

df[, 2:3] <- apply(df[, 2:3], 2, function(x) { 
    gsub("\\'", "", x)
    })