grain of table similar state_id | county_id | city_id | eventtype. eventtype binary; equal 1 or 2. want group 3 columns , see aggregate sums of eventtypes equals 1 , 2. how this?
when inner join of
select * (state_id, county_id, city_id, sum(eventtype) views select poptable eventtype = 1 group state_id, county_id, city_id) l inner join (state_id, county_id, city_id, sum(eventtype) passes select poptable eventtype = 2 group state_id, county_id, city_id) r on l.state_id = r.state_id , l.county_id = r.county_id , l.city_id = r.city_id i 500 rows. if full outer join, 3000 rows. understand there missing combinations, how them appear together?
i think want conditional aggregation:
select state_id, county_id, city_id, sum(case when eventtype = 1 1 else 0 end) views_1, sum(case when eventtype = 2 1 else 0 end) views_2 poptable group state_id, county_id, city_id; i'm not sure why doing sum(eventtype). seems strange want "1" summed in first case , "2" in second.
Comments
Post a Comment