Query to get multiple row values into multiple columns

Question

I have a CSV file with 2 columns .

Empid | SID 
:-----|-----:
12312 | S-1-5-21-3751615294 
12312 | S-1-5-21-3751615298  
12312 | S-1-5-21-3751615292 
12313 | S-1-5-21-3751615294-5078 
13546 | S-1-5-21-3751615294-50725
12312 | S-1-5-21-3751615291
14151 | S-1-5-21-3751615294-50722

For an Empid there are multiple SIDs available .I need help writing a sql SELECT query that can map(and store) these SIDs(sorted) into multiple columns.

Desired SQL Select output is below :-

+--------+---------------------------+---------------------+--------------------+--------------------+
| Empid  |    SID1                   |        SID2         |        SID3        |        SID4        |
+--------+---------------------------+---------------------+--------------------+--------------------+
| 12312  | S-1-5-21-3751-65291       | S-1-5-21-375165292  | S-1-5-21-375165294 | S-1-5-21-375165298 |
| 12313  | S-1-5-21-3751615294-5078  | NULL                | NULL               | NULL               |
| 13546  | S-1-5-21-3751615294-50725 | NULL                | NULL               | NULL               |
+--------+---------------------------+---------------------+--------------------+--------------------+

I am collecting an employee record in my application collector(using sql select queries) from a CSV file and need to collect these SIDs in his record .Maximum 4 SIDs can be possible so I created 4 attributes for SIDs .

Thanks in advance .

Why do you want to store them like that; why not have a table that maps SIDs to IDs, so you have one pair per row? — Alex Poole
– Alex Poole, Commented Dec 30, 2016 at 14:49

Gordon Linoff · Accepted Answer · 2016-12-30 14:47:36Z

2

Presumably, you know the number of columns. If so, you can do it using conditional aggregation and row_number():

select empid,
       max(case when seqnum = 1 then sid end) as sid_1,
       max(case when seqnum = 2 then sid end) as sid_2,
       max(case when seqnum = 3 then sid end) as sid_3,
       max(case when seqnum = 4 then sid end) as sid_4
from (select t.*, row_number() over (partition by empid order by empid) as seqnum
      from t
     ) t
group by empid;

If you don't know the number, then perhaps a comma-delimited list will do:

select empid, listagg(sid, ',') within group (order by sid) as sids
from t
group by empid;

A SQL query has a fixed number of columns, so a result set that has a flexible number of columns would require dynamic SQL.

answered Dec 30, 2016 at 14:47

Gordon Linoff

1.3m62 gold badges706 silver badges857 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Gurwinder Singh · Accepted Answer · 2016-12-30 14:49:27Z

2

You can use window function row_number() to assign row number to sid within each empid and then use conditional aggregation to get the final results.

select
    empid,
    min(case when rn = 1 then sid end) sid1,
    min(case when rn = 2 then sid end) sid2,
    min(case when rn = 3 then sid end) sid3,
    min(case when rn = 4 then sid end) sid4
from (select
    t.*,
    row_number() over (partition by empid order by sid) rn
from table t) group by empid;

answered Dec 30, 2016 at 14:49

Gurwinder Singh

39.7k6 gold badges62 silver badges87 bronze badges

Collectives™ on Stack Overflow

Query to get multiple row values into multiple columns

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related