R: Extract unique values in one column grouped by values in another column

Question

I had thought I could solve this with data.table but looks like it is a bit more difficult. My dataframe looks like this:

userID1   A
userID1   A
userID1   B
userID2   A
userID2   A

The output is supposed to look like this:

userID1   A
userID1   B
userID2   A

Basically create a row for each unique user and item occurence. Most of the examples are about counting unique elements but not about actually extracting these. Any suggestions?

See: https://stackoverflow.com/questions/9944816/unique-on-a-dataframe-with-only-selected-columns — George Wood, Jul 04 '17 at 16:49
Possible duplicate of [Unique on a dataframe with only selected columns](https://stackoverflow.com/questions/9944816/unique-on-a-dataframe-with-only-selected-columns) — luchonacho, Jul 04 '17 at 18:33

score 3 · Accepted Answer · answered Jul 04 '17 at 16:51

3

This is a dplyr solution that should get those results

library(dplyr)

df <- data.frame(user = c(rep("userID1",3), rep("userID2",2)), 
                 group = c("A","A","B","A","A"), 
                 stringsAsFactors = FALSE)


df <- df %>% 
      group_by(user, group) %>% 
      filter(row_number() == 1)

answered Jul 04 '17 at 16:51

Matt Jewett

3,249
1
14
21

In this instance, it may be more convenient to use: `df %>% distinct()` – George Wood Jul 04 '17 at 16:52
2

That would work too, assuming that there are truly only two columns in the original dataset, If there are really only the two columns in the dataset, then using `df[!duplicated(df),]` from base R would work as well. – Matt Jewett Jul 04 '17 at 16:56
1

Indeed. If there are more than two columns: `df %>% distinct(user, group, .keep_all = TRUE)` – George Wood Jul 04 '17 at 16:58

R: Extract unique values in one column grouped by values in another column

1 Answers1

Linked