How to do the Random sampling of of a dataset in R having the transaction ID should be together -

my sample data set following

transactionid   desc 1   1   d 1   2   c 2   d 3   l 3   g 3   h 5   h 5   b 5   h 5   f 6   d 7   f 7   v 7   f 8   f 8   d 

the sampling result should

1   1   d 1   2   c 2   d 3   l 3   g 3   h 

or

5   h 5   b 5   h 5   f 6   d 7   f 7   v 7   f 8   f 8   d 

the exact sample values not important , can important factor have keep same transaction id should in 1 sample. how can ?

you can try

 n <- 2  df[with(df, transactionid %in%           sample(unique(transactionid),n, replace=false)),]  #      transactionid desc  #1              1     #2              1    d  #3              1     #17             8    f  #18             8    d 

data

 df <- structure(list(transactionid = c(1l, 1l, 1l, 2l, 2l, 3l, 3l,   3l, 5l, 5l, 5l, 5l, 6l, 7l, 7l, 7l, 8l, 8l), desc = c("a", "d",   "a", "c", "d", "l", "g", "h", "h", "b", "h", "f", "d", "f", "v",   "f", "f", "d")), .names = c("transactionid", "desc"), class = "data.frame",  row.names = c(na,-18l))