using R - delete rows when a value repeated less than 3 times

Question

Hi I have a data.frame with 10 rows and 3 columns

    a   b c
1   1 201 1
2   2 202 1
3   3 203 1
4   4 204 1
5   5 205 4
6   6 206 5
7   7 207 4
8   8 208 4
9   9 209 8
10 10 210 5

I want to delete all rows where the same value in the column "c" repeated less than 3 times. In this example I want to remove rows 6, 9 and 10. (my real data.frame has 5000 rows and 25 cols) I tried to do it using the function rle, but I keep getting the wrong solution. any help? thanks!

Answer 1

Correct me if I'm wrong, but it seems like you want all the rows where the value in column c occurs more than twice. "Repeated" makes me think that they need to occur consecutively, which is what rle is for, but you would only want rows 1-4 if that was what you were trying to do.

That said, the code below finds the rows where the value in column c occurs more than 2 times. I'm sure this can be done more elegantly, but it works.

lines <-
"a   b c
1 201 1
2 202 1
3 203 1
4 204 1
5 205 4
6 206 5
7 207 4
8 208 4
9 209 8
10 210 5"
Data <- read.table(con <- textConnection(lines), header=TRUE); close(con)
cVals <- data.frame(table(Data$c))
Rows <- Data$c %in% cVals[cVals$Freq > 2,1]
Data[Rows,]
#  a   b c
#1 1 201 1
#2 2 202 1
#3 3 203 1
#4 4 204 1
#5 5 205 4
#7 7 207 4
#8 8 208 4

Answer 2

Building on Joshua's answer:

Data[Data$c %in% names(which(table(Data$c) > 2)), ]

Answer 3

Using unsplit is probably the easiest way to project a grouped aggregate (in this case using table to get counts, but see tapply for the general case) out to the original data.

subset(Data, with(Data, unsplit(table(c), c)) >= 3)

Equivalently and more similar to Erik's:

Data[unsplit(table(Data$c), Data$c) >= 3, ]