Question

Simplified, but for a table like:

 id time distance price
 1  20   500      8 
 2  50   500      10 
 3  90   500      12 
 4  80   1000     17 
 5  170  1000     11 
 6  180  1000     13 
 7  19   800      12 

I want to get the rows with the quickest time for the distances 500 and 1000, i.e.

 id time distance price
 1  20   500      8 
 4  80   1000     17 

If I do

select min(time) from table

that works fine for finding the price, but I can’t get the id and price – only the max/min/average/first value of all ids/prices.

I can do it with multiple look ups – e.g.

select * from table where distance = 500 and time = 20 
select * from table where distance = 1000 and time = 80 

but is there a better way that doesn’t involve 1 + (number of distances) queries (or at least provides one resultset, even if internally it uses that number of queries)

Answer 1

You will need to use an inner select:

SELECT t.id, t.time, t.distance, t.price
FROM table t
JOIN (SELECT MIN(time) as min_time, distance
        FROM table
        GROUP BY distance) as tmp
      ON (t.distance = tmp.distance AND t.time = tmp.min_time)
WHERE t.distance IN (500, 1000)