Electrino Electrino - 1 month ago 11
R Question

Extracting unique values from data frame using R

I have a data frame with multiple columns and I want to be able to isolate two of the columns and get the total amount of unique values... here's an example of what I mean:

Lets say i have a data frame df:

df<- data.frame(v1 = c(1, 2, 3, 2, "a"), v2 = c("a", 2 ,"b","b", 4))
df

v1 v2
1 1 a
2 2 2
3 3 b
4 2 b
5 a 4


Now what Im trying to do is extract just the unique values over the two columns. So if i just used unique() for each column the out put would look like this:

> unique(df[,1])
[1] 1 2 3 a
> unique(df[,2])
[1] a 2 b 4


But this is no good as it only finds the unique values per column, whereas I need the total amount of unique values over the two columns! For instance, 'a' is repeated in both columns, but I only want it counted once. For an example output of what I need; imagine the columns V1 and V2 are placed on top of each other like so:

V1_V2
1 1
2 2
3 3
4 2
5 a
6 a
7 2
8 b
9 b
10 4


The unique values of V1_V2 would be:

V1_V2
1 1
2 2
3 3
5 a
8 b
10 4


Then I could just count the rows using nrow().
Any ideas how I'd achieve this?

Answer

This is well suited for union:

data.frame(V1_V2=union(df$v1, df$v2))

#  V1_V2
#1     1
#2     2
#3     3
#4     a
#5     b
#6     4