SQL query with avg and group by

theFen picture theFen · May 22, 2012 · Viewed 125.9k times · Source

I have some problems with writing a SQL query for MySQL. I have a table with the following structure:

mysql> select id, pass, val from data_r1 limit 10;
+------------+--------------+----------------+
| id         | pass         | val            |
+------------+--------------+----------------+
| DA02959106 | 5.0000000000 |  44.4007000000 |
| 08A5969201 | 1.0000000000 | 182.4100000000 |
| 08A5969201 | 2.0000000000 | 138.7880000000 |
| DA02882103 | 5.0000000000 |  44.7265000000 |
| DA02959106 | 1.0000000000 | 186.1470000000 |
| DA02959106 | 2.0000000000 | 148.2660000000 |
| DA02959106 | 3.0000000000 | 111.9050000000 |
| DA02959106 | 4.0000000000 |  76.1485000000 |
| DA02959106 | 5.0000000000 |  44.4007000000 |
| DA02959106 | 4.0000000000 |  76.6485000000 |

I want to create a query that extracts the following information from the table:

id, AVG of 'val' for 'pass' = 1, AVG of 'val' for 'pass' = 2, etc

The result of the query should look like this:

+------------+---------+---------+---------+---------+---------+---------+---------+
| id         | val_1   | val_2   | val_3   | val_4   | val_5   | val_6   | val_7   |
+------------+---------+---------+---------+---------+---------+---------+---------+
| DA02959106 | 186.147 | 148.266 | 111.905 | 76.3985 | 44.4007 | 0       | 0       |
+------------+---------+---------+---------+---------+---------+---------+---------+

with more rows for each unique 'id', of course.

I already tried some queries like

SELECT id, pass, AVG(val) AS val_1 FROM data_r1 WHERE pass = 1 GROUP BY id;

This returns the correct result, but I have to expand it with results for the other possible values of 'pass' (up to 7)

I tried to use a nested SELECT within AVG but this didn't work because I didn't figure out how to correctly limit it to the current 'id'.

I then created Views to represent the result of each query for 'pass' = 1, 'pass' = 2, etc. But for most ids the highest value for 'pass' is 5. When using JOIN queries to get the final result from the views I received an empty result set, because some of the Views are empty / don't have values for a specific 'id'.

Any ideas?

Answer

Marco picture Marco · May 22, 2012

If I understand what you need, try this:

SELECT id, pass, AVG(val) AS val_1 
FROM data_r1 
GROUP BY id, pass;

Or, if you want just one row for every id, this:

SELECT d1.id,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 1) as val_1,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 2) as val_2,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 3) as val_3,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 4) as val_4,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 5) as val_5,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 6) as val_6,
    (SELECT IFNULL(ROUND(AVG(d2.val), 4) ,0) FROM data_r1 d2 
     WHERE d2.id = d1.id AND pass = 7) as val_7
from data_r1 d1
GROUP BY d1.id