I'm using aggregate to get the mean of a value as determined by three other values. I'd also like to get a count of how many values met those three criteria.
For example, I have df:
df <- data.frame(id = c(1,1,1,2,2,2,3,3,3,3),
col1=sample(1:10, 10, replace = TRUE),
col2 = c(1,2,1,2,1,1,2,2,2,1),
col3 = c("a","b","c", "b","a","a","b","a","c","a"),
col4 = c("yes","no","no", "no","yes","yes","yes","no","no","yes"))
And I run aggregate to get the mean for each unique occurence, like:
df_agg <- aggregate(col1~col2+col3+col4, df, FUN = mean)
What I want to know is that in addition to the mean of rows that have col2 =1, col3=a, col4=yes that there are also 3 unique id values that meet the criteria. Basically a sample size or n number, but of the unique values. I tried doing df_agg <- aggregate(id~col2+col3+col4, df, FUN = length), but this gives me the total count of id rows, I just want the unique ones.
Something like this, an example of one row from my sample data that would have more than 1 unique id.
An example output for a line that would have more than unique ID:
col2 col3 col4 mean count
|---------|--------|---------|--------|--------|
| 1 | a | yes | 5 | 3 |
|---------|--------|---------|--------|--------|
