Select query that eliminate duplicates

Question

I am struck at the below scenario. need your help in modify my query based on the below sample data. My objective is to create a EVENT AND EVENT_LOW using the data available from TEM_GT and TXN table.
TEM_GT (global temporary table)
est_id primary key, trans_id, trp_id, amount
1 111 2221 1.5
2 111 3332 2.0
3 112 4443 3.0
TXN table
trans_id, trans_type
111 type1
112 type1
EVENT table
event_id primary key, trans_id, trans_type, flag.
1000 111 type1 N
1001 112 type1 N
EVENT_LOW table
event_low_id primary key, event_id, est_id, amount.
9991 1000 1 1.5
9992 1000 1 2.0
9993 1001 2 3.0
insert into(event_low_id, event_id, est_id, amount)
select event_low_id_s.nextval e.event_id, tg.est_id, tg.amount from
from TEM_GT tg, EVENT e
WHERE
tg.trans_id = e.trans_id
AND e.flag = 'N'
Based on TEM_GT and TXN gt, populating data into EVENT table. Now when I try to populate data in the EVENT_LOW table, the above query returns 5 rows instead of 3 rows. This is due to trp_id in TEM_GT table. I do not want to add trp_id in EVENT table, and want to handle the duplicate elimination in the above select query.
I wrote the below sql. It almost resolve my problem. But sometimes it returns different value for the amount column. I wanted my event_low result as below for the est_id 1.
event_low_id primary key, event_id, est_id, amount.
9991 1000 1 1.5
9992 1000 1 2.0
But sometime it returns as below
event_low_id primary key, event_id, est_id, amount.
9991 1000 1 1.5
9992 1000 1 1.5
or
event_low_id primary key, event_id, est_id, amount.
9991 1000 1 2.0
9992 1000 1 2.0
select *
from (select x.*,
row_number() over (partition by event_id order by event_id) rn
from (seelct e.event_id, tg.est_id, tg.amount
from TEM_GT tg, EVENT e
WHERE
tg.trans_id = e.trans_id
AND e.flag = 'N'
) x
where rn = 1
I am using Oracle 11g. Any help. Thanks in advance.

Answer 1

Hi Suresh,
Please read SQL and PL/SQL FAQ
Whenever you create a thread always post sample data as CREATE TABLE and INSERT statements.
Additionally when you put some code or output please enclose it between two lines starting with {noformat}{noformat}
i.e.:
{noformat}{noformat}
SELECT ...
{noformat}{noformat}
Regarding your problem you said:
Based on TEM_GT and TXN gt, populating data into EVENT table. Now when I try to populate data in the EVENT_LOW table, the above query returns 5 rows instead of 3 rows.
The above query cannot return 5 rows with the data you have posted. You have 3 rows in TEM_GT and 2 rows in EVENT table. The table are joined ontg.trans_id = e.trans_id AND e.flag = 'N'
And the result of the query with your posted data is *3 rows*. So try to be clear.
If your intention is to insert into EVENT_LOW only one row from TEM_GT for the same trans_id then you have to tell us which is the criteria to choose the row between the duplicates (the first by est_id order??).
Also be careful to post valid statements. The statement below is not valid as it is missing the table name after INSERT INTO and there are 2 consecutive FROM keywords:insert into(event_low_id, event_id, est_id, amount)
select event_low_id_s.nextval e.event_id, tg.est_id, tg.amount from
from TEM_GT tg, EVENT e
WHERE
tg.trans_id = e.trans_id
AND e.flag = 'N'
I will make the assumption that you want to select the first row order by est_id during insertion. If this is not what you want, please change the analytic function:INSERT INTO event_low (event_low_id, event_id, est_id, amount)
WITH got_data AS
SELECT e.event_id, tg.est_id, tg.amount
, ROW_NUMBER() OVER(PARTITION BY tg.trans_id ORDER BY tg.est_id) AS rn
FROM tem_gt tg, event e
WHERE tg.trans_id = e.trans_id
AND e.flag = 'N'
SELECT event_low_id_s.NEXTVAL
, event_id, est_id, amount
FROM got_data
WHERE rn = 1;
If this is not what you want please post CREATE TABLE and INSERT statement, and post your expected output (for insertion into EVENT_LOW table) for the data you have posted.
Regards.
Al