Selecting rows from a data frame

Question

Please consider this example code:

d1 <- c(1,2,2,3,4,3)
d2 <- c(10,11,12,13,14,15)

dt <- data.frame(d1,d2)

sample.index <- c(2,3)

dt[dt$d1 %in% sample.index, ]

This returns

  d1 d2
2  2 11
3  2 12
4  3 13
6  3 15

which is OK. However, if we have

sample.index <- c(2,2,3)

then the code still returns the same result. Instead I want the rows matching 2 to be returned twice because 2 appears twice in sample.index - how can I achieve this ?

Answer 1

Maybe this:

sample.index <- c(2,2,3)
merge(dt,data.frame(d1 = sample.index))
  d1 d2
1  2 11
2  2 11
3  2 12
4  2 12
5  3 13
6  3 15

Answer 2

This is begging for some data.table syntax sugar (goes without mention that it will also be faster):

library(data.table)

d1 <- c(1,2,2,3,4,3)
d2 <- c(10,11,12,13,14,15)

# Note, I set the key to d1
dt <- data.table(d1, d2, key = 'd1')

dt[J(c(2,3))]
#   d1 d2
#1:  2 11
#2:  2 12
#3:  3 13
#4:  3 15

dt[J(c(2,2,3))]
#   d1 d2
#1:  2 11
#2:  2 12
#3:  2 11
#4:  2 12
#5:  3 13
#6:  3 15

Also note that data.table join and merge.data.frame result in somewhat different final ordering.