asked on

Select distinct column

Hi,
I have a database with where a value in a column may appear many times. It would be easy to select the distinct values from this column on its own but I need to also return other columns. Therefore the distinct clause will apply to all the columns being returned (so distinct is where column a, b, c and d are distinct). I suppose then my question is simply how do I return distinct values from column a and also return column b and c (therefore just taking say the distinct values from column a and ignoring the fact that value may appear again in the column).
Hope you can understand me.
thanks in advance

dave_gr

select distinct colA, colB, colC, colD from sourceTable

or

select colA, colB, colC, colD
from sourceTable
group by colA, colB, colC, colD

Not sure if that answers your question mind, as I'm not really sure what you are asking.

David

manukaucouncil

ASKER

when you use "select distinct colA, colB, colC, colD from sourceTable" does that mean that colA, colB, colC and colD need to be distinct together or is it just distinct colA? Say for example I want to just return distinct colB values... how would I do this as select colA, distinct colB....... does not work.

dave_gr

If you just want distinct colB values then you can do

select distinct colB from sourceTable

for distinct colA then

select distinct colA from sourceTable

for every distinct combination of colA and colB you

select distinct colA, colB from sourceTable

What are your columns? What are you trying to achieve?

manukaucouncil

ASKER

This is my script:
select amount_id, adjust_id, amttyp_id, description, order_no, essbase_account_id, essbase_entity_id, ltrim(STR(adjust_id)) + 'a' + ltrim(STR(order_no)) AS adjust_order, amount_id from atm_amount where essbase_account_id in (select dimension_1_node_key from dss_dimension_1 where node_desc = 'PS3001') and adjust_id in (select distinct adjust_id from atm_adjustment where adjtyp_id in (1,2) and status_id <7) order by adjust_id,order_no

Bascically I just want to return the distinct adjust_id's in the outer select statement but with the other information still attached to it.

appari

do you mean you want
adjust_id in (select distinct adjust_id from atm_adjustment where adjtyp_id in (1,2) and status_id <7) order by adjust_id,order_no
to
adjust_id in (select distinct adjust_id, other1,other2 from atm_adjustment where adjtyp_id in (1,2) and status_id <7) order by adjust_id,order_no

if you add other information in the outer select you cannot use it with "in" condition.

why do you need other information in that outer sql? can you explain what you are trying to do?

manukaucouncil

ASKER

Say for example I have this data:

ClassName ClassDescription StudentName Mark
SQL Server All about SQL Jeremy A
VB All about VB Jeremy A+
SQL Server All about SQL 2 Bob B

Now I want to do something like this: "select distinct classname, classdescription, studentname, mark from table"
The problem with this script is that it will recognise rows 1 and 3 as distinct because they have a different description.
So I really only want the distinct clause to apply to classname so the results return as follows:

ClassName ClassDescription StudentName Mark
SQL Server All about SQL Jeremy A
VB All about VB Jeremy A+

So it just ignores the second occurance of SQL Server in classname.
I hope that makes it a bit clearer.

ASKER CERTIFIED SOLUTION

appari

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

manukaucouncil

ASKER

Ahhh that certainly makes sense. Thanks