Partially distinct select

Question

I have a table with fields utm and tracking_id. There may be many tracking_ids for each utm. So I need to select distinct utm and any one tracking id for each utm (let`s say thefirst one).

For example, we got the table:

utm   | tracking_id
-------------------
ddff  |      1
ddff  |      2
llzz  |      3
ddff  |      4
ohoh  |      5
ohoh  |      6

And as an output i want to get:

utm   | tracking_id
-------------------
ddff  |      1
llzz  |      3
ohoh  |      5

I use PostgreSQL 9.1.
Is there a way to do it with SQL?

so you want the lowest tracking_id for each group of utm entities, right? — STT LCU, Sep 25 '13 at 11:30

score 5 · Answer 1 · answered Sep 25 '13 at 11:28

5

select utm, min(tracking_id)
from t
group by utm

answered Sep 25 '13 at 11:28

Clodoaldo Neto

108,856
25
211
247

Tracking_id is not present in the group by clause. This is a not valid SQL statement. – AlexVPerl Dec 15 '15 at 05:16

score 2 · Accepted Answer · answered Sep 25 '13 at 11:37

if you have only one column, than simple aggregate is what you want, go with Clodoaldo Neto's advice. If you have more than one columns, you can use dictinst on syntax:

select distinct on(utm)
    utm, tracking_id, column1, column2, ...
from t
order by utm, tracking_id

sql fiddle demo

Partially distinct select

2 Answers2