Question

I’m trying to build a SQL statement that will select the highest id but only the top value if duplicates exist.

id – value

• 1491825 – 3339
• 1491824 – 3170
• 1491823 – 3143
• 1491822 – 3422
• 1491821 – 3422
• 1491820 – 3422
• 1491819 – 3422
• 1491818 – 3422
• 1491817 – 3261
• 1491816 – 3170
• 1491815 – 3170
• 1491814 – 3170
• 1491813 – 3170
• 1491812 – 3170
• 1491811 – 3170
• 1491810 – 3170
• 1491809 – 3170
• 1491808 – 3170
• 1491807 – 3251
• 1491806 – 3422

Expected Output:

id – value

• 1491825 – 3339
• 1491824 – 3170
• 1491823 – 3143
• 1491822 – 3422
• 1491817 – 3261
• 1491807 – 3251

Answer 1

select max(id),value From foo
group by value
order by max(id) desc

See: http://www.sqlfiddle.com/#!3/4ee34/3

Edit. Darn. Beaten to it by astander.