CODEWITHSUNDEEP
rtime
Megan
Megan

Reputation: 195

Group consecutive times together within a threshold to get a TimeGroup output in R

I have a column or vector of times. The data were originally evenly spaced at 2 min intervals but I subsetted the data to look at records of interest. I now want to group all time records together that are consecutively 2 mins apart. The result would be a column/vector called 'TimeGroup' to identify which records are grouped together.

DateTime<-as.POSIXct( strptime(c("2013-05-27 12:41:52 UTC", "2013-05-27 13:25:52 UTC", "2013-05-27 13:27:52 UTC",
  "2013-05-27 13:29:52 UTC" ,"2013-05-27 13:31:52 UTC" ,"2013-05-27 13:33:52 UTC",
  "2013-05-27 13:35:52 UTC" ,"2013-05-27 13:37:52 UTC" ,"2013-05-27 13:39:52 UTC",
  "2013-05-27 13:41:52 UTC", "2013-05-27 13:43:52 UTC", "2013-05-27 13:45:52 UTC",
  "2013-05-27 13:47:52 UTC" ,"2013-05-27 13:49:52 UTC" ,"2013-05-27 13:51:52 UTC",
  "2013-05-27 13:53:52 UTC", "2013-05-27 13:55:52 UTC", "2013-05-27 13:57:52 UTC",
  "2013-05-27 13:59:52 UTC", "2013-05-27 14:01:52 UTC" ,"2013-05-27 14:03:52 UTC",
  "2013-05-27 14:05:52 UTC", "2013-05-27 14:07:52 UTC", "2013-05-27 14:41:52 UTC",
  "2013-05-27 14:43:52 UTC" ,"2013-05-27 14:45:52 UTC", "2013-05-27 14:47:52 UTC",
  "2013-05-27 15:11:52 UTC" ,"2013-05-27 15:13:52 UTC", "2013-05-27 15:15:52 UTC",
  "2013-05-27 15:17:52 UTC" ,"2013-05-27 15:21:52 UTC", "2013-05-27 15:23:52 UTC",
  "2013-05-27 15:25:52 UTC" ,"2013-05-27 15:27:52 UTC", "2013-05-27 15:29:52 UTC",
  "2013-05-27 15:33:52 UTC" ,"2013-05-27 15:35:52 UTC" ,"2013-05-27 15:37:52 UTC",
  "2013-05-27 15:39:52 UTC", "2013-05-27 15:41:52 UTC" ,"2013-05-27 15:43:52 UTC",
  "2013-05-27 15:45:52 UTC" ,"2013-05-27 15:47:52 UTC" ,"2013-05-27 16:03:52 UTC",
  "2013-05-27 16:05:52 UTC", "2013-05-27 16:07:52 UTC", "2013-05-27 16:09:52 UTC",
  "2013-05-27 16:11:52 UTC", "2013-05-27 16:13:52 UTC", "2013-05-27 16:15:52 UTC",
  "2013-05-27 16:17:52 UTC" ,"2013-05-27 16:19:52 UTC" ,"2013-05-27 16:21:52 UTC",
  "2013-05-27 16:23:52 UTC", "2013-05-27 16:25:52 UTC" ,"2013-05-27 16:27:52 UTC",
  "2013-05-27 16:29:52 UTC" ,"2013-05-27 16:31:52 UTC" ,"2013-05-27 16:33:52 UTC",
  "2013-05-27 16:39:52 UTC" ,"2013-05-27 16:41:52 UTC" ,"2013-05-27 16:43:52 UTC",
  "2013-05-27 16:45:52 UTC" ,"2013-05-27 16:47:52 UTC" ,"2013-05-27 16:49:52 UTC",
  "2013-05-27 16:51:52 UTC" ,"2013-05-27 16:53:52 UTC" ,"2013-05-27 16:55:52 UTC",
  "2013-05-27 16:57:52 UTC" ,"2013-05-27 16:59:52 UTC" ,"2013-05-27 17:01:52 UTC",
  "2013-05-27 17:03:52 UTC" ,"2013-05-27 17:05:52 UTC" ,"2013-05-27 17:07:52 UTC",
  "2013-05-27 17:09:52 UTC", "2013-05-27 17:11:52 UTC" ,"2013-05-27 17:13:52 UTC",
  "2013-05-27 17:15:52 UTC", "2013-05-27 17:17:52 UTC" ,"2013-05-27 17:19:52 UTC"), format="%Y-%m-%d %H:%M:%S"),1970-01-01, tz="UTC")


diff(DateTime)
Time differences in mins
 [1] 44  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2 34  2  2  2 24  2  2  2  4  2  2  2
[35]  2  4  2  2  2  2  2  2  2 16  2  2  2  2  2  2  2  2  2  2  2  2  2  2  2  6  2  2  2  2  2  2  2  2
[69]  2  2  2  2  2  2  2  2  2  2  2  2

From this example data, there would be 7 TimeGroups called 1,2,3,4,5,6,7. The records not apart of a time group would be NA.

Upvotes: 0

Views: 32

Answers (1)

Wimpel
Wimpel

Reputation: 27782

try the following

library( data.table )
DT <- data.table( DateTime = DateTime )
DT[, group := c(0,cumsum( diff(DateTime) > 2)) ]

Upvotes: 1

Related Questions