SQL选择唯一实体

Question

I have in my table:

ID Name   Age
1  James  15
2  James  16
3  Joseph 16
4  Joseph 18
5  Steve  4

and I want to return:

5 Steve 4
1 James 15
3 Joseph 16

ie. Unique entries based on the Name row (choose the one with lowest ID when there are multiple of the same Name), and ordered by Age

What would be the correct SQL statement?

I have:

Select * FROM table Where True GROUP BY Name ORDER BY Age

It seems to be returning a lot of unique results, but it's too slow (table has >250,000 entries and it takes >30seconds to process), I wonder if there is a faster way to do this?

Answer 1

try this

SELECT *
FROM table WHERE ID in (SELECT MIN(ID) FROM table group by name)
ORDER by age

You just group the datasets by name and get MIN(ID) and use this as 'where' condition for your selection

Answer 2

Another way of querying this, usually better where performance is needed, provided you have (and that's NOT an option here) an INDEX on Name :

SELECT *
FROM your_table t1
LEFT JOIN your_table t2 ON t1.Name = t2.Name AND t2.ID < t1.ID
WHERE t2.ID IS NULL;

A quick explanation here .

Answer 3

Since you want to select based on the minimum ID , which is a unique field, you just need to get the minimum ID s per name first, then join those ID s to the main table. This will be much faster than using a subquery in a WHERE as the subselect of the minimums gets executed just one time, whereas a subquery would execute for each row in the main table (250,000+ times!).

SELECT
    b.*
FROM
    (
        SELECT MIN(ID) AS minID
        FROM tbl
        GROUP BY Name
    ) a
INNER JOIN
    tbl b ON a.minID = b.ID
ORDER BY
    b.Age