How to combine row with some condition?

Question

I have dataset like this,

> data
    ID    AccessTime   ReferrerCode
    101   01:17:40     910
    103   01:27:53     135
    103   01:33:12     222
    202   02:04:11     921
    202   03:40:30     106 
    103   03:45:02     734

And I want to transform my dataset like this,

> data
ID   ReferrerCode
101   910
103   135 222
202   921
202   106
103   734

My condition is ReferrerCode will be combined if the data has the same ID and the same hour of AccessTime. What should I do to make it happen?

Answer 1

Here's a base R solution with aggregate() , and sub() to extract the hour from AccessTime .

aggregate(ReferrerCode ~ ID + sub(':.*$', '', AccessTime), c, data=data)[,-2]
#   ID ReferrerCode
#1 101          910
#2 103     135, 222
#3 202          921
#4 103          734
#5 202          106

Answer 2

If you want to get the result as character in ReferrerCode use this:

aggregate(ReferrerCode ~ ID + hour(AccessTime), data, FUN=function(x) paste(x, collapse=" "))[,c(1,3)]
       ID ReferrerCode
1 101          910
2 103      135 222
3 202          921
4 103          734
5 202          106

I assumed that AccessTime is in POSIXlt format. From character you can convert by

data$AccessTime <- as.POSIXlt(data$AccessTime, format="%H:%M:%S")