Question

Using a h2 database I want to get average values of a column while grouping entries based on a timestamp.

My table looks like this:

TIME                    SID             SERVICE         TET     
2012-11-21 11:31:45.433 164DSQL0L4JW7   service1        3868    
2012-11-21 11:31:45.608 164DSS6TJS321   service2        350     
2012-11-21 11:31:45.711 164DSROS9DNXA   service2        285     
2012-11-21 11:31:45.766 164DSH2PMPY4M   service1        4044    
2012-11-21 11:31:45.778 164DSOGRCKG8J   service1        4248    
2012-11-21 11:31:45.82  164DSRMV0QMFT   service1        3766    
2012-11-21 11:31:46.089 164DSP1L9W6MC   service1        4257    
2012-11-21 11:31:46.226 164DSOPUNUFJF   service3        703     
2012-11-21 11:31:46.752 164DSLJHST6E8   service3        411     
2012-11-21 11:31:46.781 164DSHPI1T3J9   service3        393     
2012-11-21 11:31:45.24  164DSQI4RY330   service2        235     
2012-11-21 11:31:45.268 164DSPRDHROQN   service2        197     
2012-11-21 11:31:45.991 164DSRUUGQ8LN   service2        380     
2012-11-21 11:31:46.338 164DSMARDXMD3   service3        783  

I’d like to extract average values for the column TET grouped by the column TET.
I want to have the interval flexible (e.g. 15 seconds and for another query 5 minutes).

I found this question which is pretty much what I need, but unfortunately I can’t get any of the solutions to work in H2 (there’s no strftime).

Answer 1

H2 supports a number of date and time functions, for example DATEDIFF and DATEADD. Example:

drop table if exists test;
create table test(id int primary key, ts timestamp);
insert into test 
select x, dateadd(s, x, now()) from system_range(1, 100);
select  dateadd(s, datediff(s, now(), ts) / 10 * 10, now()) as diff,
avg(id) from test group by diff order by diff;