Reputation: 180
So i have been scratching my head over this one,mostly because i am on access 2010 and most of the queries i have found on the internet have commands that do not work on access.
id name date qty created
====================================================
1 abc 01/2016 20 06/07/2016 11:00
2 abc 02/2016 20 06/07/2016 11:00
3 abc 03/2016 20 06/07/2016 11:00
4 abc 01/2016 30 06/07/2016 13:00
I need to pull out a recordset like this:
id name date qty created
====================================================
2 abc 02/2016 20 06/07/2016 11:00
3 abc 03/2016 20 06/07/2016 11:00
4 abc 01/2016 30 06/07/2016 13:00
the created field is just a timestamp, the date field is a "due date". basically i need to pull out the most recent qty for each name and date. the ID is unique so i can use it instead,if its easier. By far i've got:
SELECT m1.date, m1.name, m1.created
FROM table AS m1 LEFT JOIN table AS m2 ON (m1.created < m2.created) AND
(m1.date = m2.date)
WHERE m2.created IS NULL;
but this one gives me only the most recent conflicted data, ie. the record n°4 in my example.i also need the other two records. any thoughts?
Upvotes: 3
Views: 4390
Reputation: 72225
Try this:
SELECT m1.date, m1.name, m1.qty, m1.created
FROM table AS m1
JOIN (
SELECT date, name, MAX(created) AS created
FROM table
GROUP BY date, name
) AS m2 ON m1.date = m2.date AND m1.name = m2.name AND m1.created = m2.created
Upvotes: 0
Reputation: 40491
Try using NOT EXISTS()
:
SELECT * FROM YourTable t
WHERE NOT EXISTS(SELECT 1 FROM YourTable s
WHERE t.date = s.date and s.created > t.created
AND t.name = s.name)
I think you are also missing a condition so I've added it:
and t.name = s.name
You didn't tag your RDBMS, if its SQL-Server/Oracle/Postgresql you can use ROW_NUMBER()
:
SELECT s.date, s.name, s.created FROM (
SELECT t.*,
ROW_NUMBER() OVER(PARTITION BY t.date,t.name ORDER BY t.created DESC) as rnk
FROM YourTable t) s
WHERE s.rnk = 1
Upvotes: 4