Count of size range

Question

Hi
could some one assist.
Our requirement is Number of tickets by Size per each month, the data in Trade table like this
Trade_id-->Size-->trade_Date-->Market
1-------->110----->01/02/11-->AA
2-------->80------>25/01/11-->AB
3-------->210----->17/01/11-->BB
4-------->150----->15/02/11-->AA
5-------->260------>20/12/10-->AB
6-------->320------>15/12/10-->BB
7------->60------->12/12/11-->AA
8-------->60------->20/01/11->AB
Final result would be
Month-->0-100-->101-200-->201-300-->301-400
Dec-10---->1 ---------->0 ------------>1 ------>1
Jan-11------>2 --------->0 ------------>1 -------->0
Feb-11----->0----------->2-------------->0-------->0
And assist would be appreciated.
Thanks in Advance
Megna

Answer 1

I guess you want something like this. Btw. Your sample data didn't match the wanted outcome. I guess one of the values was mistyped but I keept the values as you provided them.
For future posts try to give an example using a small insert script or a with clause. THis helps us to help you. Also there is a {<b></b>code} tag that you should put before and after your code to improve readablility and to keep indentions.
with testdata as (select 1 trade_id, 110 sizecol, to_date('01/02/11','dd/mm/rr') trade_date, 'AA' market from dual union all
                  select 2 trade_id, 80 sizecol, to_date('25/01/11','dd/mm/rr') trade_date, 'AB' market from dual union all
                  select 3 trade_id, 210 sizecol, to_date('17/01/11','dd/mm/rr') trade_date, 'BB' market from dual union all
                  select 4 trade_id, 150 sizecol, to_date('15/02/11','dd/mm/rr') trade_date, 'AA' market from dual union all
                  select 5 trade_id, 260 sizecol, to_date('20/12/10','dd/mm/rr') trade_date, 'AB' market from dual union all
                  select 6 trade_id, 320 sizecol, to_date('15/12/10','dd/mm/rr') trade_date, 'BB' market from dual union all
                  select 7 trade_id, 60 sizecol, to_date('12/12/11','dd/mm/rr') trade_date, 'AA' market from dual union all
                  select 8 trade_id, 60 sizecol, to_date('20/01/11','dd/mm/rr') trade_date, 'AB' market from dual
                  /* end of test data creation */
select to_char(trunc(trade_date,'MONTH'),'Mon-YY') as "month",
       count(*) as "total",
       count(case when sizecol between 0 and 100 then 1 end) as "bucket 0-100",
       count(case when sizecol between 101 and 200 then 1 end) as "bucket 101-200",
       count(case when sizecol between 201 and 300 then 1 end) as "bucket 201-300",
       count(case when sizecol between 301 and 400 then 1 end) as "bucket 301-400",
       count(case when sizecol >= 401 then 1 end) "bucket 401+"
from testdata
group by trunc(trade_date,'MONTH')
order by trunc(trade_date,'MONTH');
month     total     bucket 0-100     bucket 101-200     bucket 201-300     bucket 301-400     bucket 401+
Dec-10     2     0     0     1     1     0
Jan-11     3     2     0     1     0     0
Feb-11     2     0     2     0     0     0
Dec-11     1     1     0     0     0     0This takes advantage of the fact that the count function will not count NULL values. The case expression will deliver NULL if the sizecol doesn't fit into the appropriate bucket.
There are other solutions that would deliver a similar result, but are more flexible. For example the NTILE or WIDTH_BUCKET function could be used. However they would return/consider the output as rows and not columns. It still might be worth exploring those solutions.
Edited by: Sven W. on Feb 17, 2011 10:56 AM