Use of DISTINCT in SELECT close

[Ivan Smahin]

Hi Stan,

> On Nov 30, 2015, at 3:03 PM, Stanley Roche Busk <maxprog at mac.com> wrote:
> 
> 
> Ok, actually ‘a' is a string and ‘b' a record number. I try to get all the different versions of ‘a’. ‘a’ may or may not be different, ‘b’ is always different.
> 
> SELECT (a), b FROM c GROUP BY a   ??? I am not in front of my working dev. computer right now so I can’t try :-(
> 
> Stan
> 
> 

François Van Lerberghe is absolutely right.
Actually, you want to group results by "a" column. It means - some single value (based on group of "b" values) for each distinct "a" value.
You should only decide what aggregation function to apply for each group.

Something like this one:

SELECT a, MIN(b) FROM c GROUP BY a;

Another example - if you want to get a list of "b" values (as single value) for each "a" group:

SELECT a, GROUP_CONCAT(b) FROM c GROUP BY a;

...

>> On Nov 30, 2015, at 1:38 PM, François Van Lerberghe <frvanlerberghe at gmail.com> wrote:
>> 
>> Hello Stan,
>> 
>> You must decide what you do for the b column when there is duplicates on column a : you take the first, the last, …
>> 
>> If you decide to take the first value, your sql query could be this one : SELECT a, FIRST(b) FROM c GROUP BY a
>> 



-- 
Best regards,
Ivan Smahin
Senior Software Engineer
Paradigma Software, Inc
Valentina - The Ultra-Fast Database
http://www.valentina-db.com