MySQL Using ORDER BY On Grouped (GROUP BY) Results

Question

I am clearly misundersatanding something here with MySQL's GROUP BY as it is changing the order of my results.

Using this example SQL data:

    CREATE TABLE IF NOT EXISTS `example_table` (
      `ID` int(11) NOT NULL AUTO_INCREMENT,
      `GROUP_NAME` text NOT NULL,
      `ORDER_COLUMN` int(11) NOT NULL,
      PRIMARY KEY (`ID`)
    ) ENGINE=InnoDB  DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci AUTO_INCREMENT=1 ;



    INSERT INTO `example_table` (`ID`, `GROUP_NAME`, `ORDER_COLUMN`) VALUES
    (NULL, '271007K240.003:10', 70),
    (NULL, '271007K240.003:10', 90),
    (NULL, '271007K240.003:10', 100),
    (NULL, '271007K240.003:10', 50),
    (NULL, '271007K240.003:10', 80),
   (NULL, '271007K240.003:10', 60);

Now I've created this data as an example so there would be lots of different values in the GROUP_NAME column., but for clairty I've just included what demonstrates the issue.

Running this very simple query, returns the highest ORDER_COLUMN (ID: 3) at the top as expected:

SELECT
    *
FROM
    `example_table`
ORDER BY
    ORDER_COLUMN
DESC

However I've actually wanting to group by the column I've named here GROUP_NAME , sowith this in mind I was anting to do asomething like so:

SELECT
    *
FROM
    `example_table`
GROUP BY
    GROUP_NAME
ORDER BY
    ORDER_COLUMN
DESC

Now doing this just simply returns the first row, and ignores the ORDER BY.

I then thought to achieve what I need, I would need to use a sub-query with the data pre-sorted and then the group by can just group the sub-queries data-set like so:

SELECT
    *
FROM
    (
    SELECT
        *
    FROM
        `example_table`
    ORDER BY
        ORDER_COLUMN
    DESC
) AS TEMP_TABLE

GROUP BY GROUP_NAME

Unfortunately though, this still returns just the first row of the table. What am I doing wrong here?

Answer 1

With this query:

select group_name, max(order_column) order_column
from example_table
group by group_name

you can get the max value of order_column for each id .
Then join it to the table:

select t.* 
from example_table t inner join ( 
  select group_name, max(order_column) order_column
  from example_table
  group by group_name
) g on g.group_name = t.group_name and g.order_column = t.order_column

See the demo .
Results:

| ID  | GROUP_NAME        | ORDER_COLUMN |
| --- | ----------------- | ------------ |
| 3   | 271007K240.003:10 | 100          |

Answer 2

Not completely sure what you want to accomplish, but did you try:

ORDER BY
    GROUP_NAME,
    ORDER_COLUMN

If that doesn't do the trick, how about adding more data with different group names and an example of what you expect the output to look like