Select * 其中 Column_A 和 Column_B 是唯一的

Question

I've got a table with about 70 million rows. I start by narrowing it down to about 250k rows by selecting from a specific sim_id.

Now, I need to select each first row that has a unique combination of sorted_hand AND ev. Put another way, I need to disregard all rows that duplicate an existing row's sorted_hand and ev and select what remains.

Table looks like this:

|| id || sim_id || sorted_hand || ev
|| 1 || 1 || 23AA || 2453
|| 2 || 1 || 23AA || 2453 
|| 3 || 1 || 23AA || 2453 
|| 4 || 1 || 2233 || -548 
|| 5 || 1 || 23AK || -296
|| 6 || 1 || 2233 || -548 

In the example selection above, I would want the resulting query to select:

|| 1 || 1 || 23AA || 2453
|| 4 || 1 || 2233 || -548
|| 5 || 1 || 23AK || -296

All of the other rows are disregarded because they have the same exact ev and`sorted_hand as another row.

Answer 1

You can grab the min(id) and just GROUP BY the rest of the columns to achieve this:

SELECT min(id) as id, sim_id, sorted_hand, ev FROM yourtable GROUP BY sim_id, sorted_hand, ev;

In essence, it's not so much "disregarding" but rather grouping records and just returning the first id encountered for that group.

Answer 2

You can filter with a correlated subquery:

select t.*
from mytable t
where t.id = (
    select min(t1.id)
    from mytable t1
    where t1.sorted_hand = t.sorted_hand and t1.ev = t.ev
)

For performance, you want an index on (sorted_hand, ev, id) .

Alternatively, If you don't care about sim , or if it is acceptable to add it to the group, then you can simply use aggregation

select min(id) id, sim, sorted_hand, ev
from mytable
group by sim, sorted_hand, ev