Try SELECT client, receive_day, receive_hour as start_time, receive_hour+1 as end_time FROM some_table WHERE client='xyz' AND receive_day=7 ORDER BY start_time;
On Mon, Oct 22, 2012 at 4:41 PM, dyuti a <hadoop.hiv...@gmail.com> wrote: > Hi all, > I have a hive table with 235 million records. > > SAMPLE INPUT: > receive_year receive_day receive_hour client > 2012 7 17 > xyz > 2012 7 02 > xyz > 2012 7 17 > btn > 2012 7 04 > snm > 2012 7 05 > btn > 2012 7 02 > snm > > receive_hour is an int field having values from 01 to 24 (transaction > received hour in 24 hour basis). > > EXPECTED OUTPUT: > > The transaction received per hour basis i.e., 01 (start_time) t0 02 > (end_time) . so that the receive_hour column should be displayed as > start_time and end_time so that every one hour transaction received for > particular client on that day. > > client receive_day start_time end_time > xyz 7 01 > 02 > xyz 7 02 > 03 > xyz 7 03 > 04 > continues up to 23 24 > > like the same for other clients too. > > am facing problem in displaying the single column values into two column in > hive. i.e., 1 - 2 , 2- 3 ............. 23 - 24. > > I have written the query to display the other details happened in > transaction but got stuck in the above scenario. > > Please let me know is this possible in hive, if so then pls guide me. > > > Thanks for your help in advance! > > > Regards, > dti > > > > >