MVSFORUMS.com :: View topic - VERTICAL to HORIZONAL COLUMNS

through

#1: VERTICAL to HORIZONAL COLUMNS Author: NASCAR9, Location: California

Posted: Tue Dec 18, 2018 2:33 pm
—
The data is in a vertical format with a key of FAMILY_ID that allows grouping a Family. The data looks something like this:

Code:

FAMILY_ID RELATION_ID FIRST_NAME LAST_NAME
123456    0       HOBO    KELLY
123456    1       BOZO    CLOWN
123456    6       YUCKO    CLOWN
123456    9       SIDESHOW BOB
654321    0       ROY RODGERS
654321    2 DALE EVANS
654321    3       CISCO    KID
654321    4       LUCAS    McCAIN
654321    9       HOPALONG    CASSIDY

This query will flatten the data. I allowed for 10 FAMILY members in a FAMILY_ID. This is just an example.

Code:

SELECT Z.FAMILY_ID,
MAX(DECODE(Z.RELATIONID_RANK, 01, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D1,
MAX(DECODE(Z.RELATIONID_RANK, 02, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D2,
MAX(DECODE(Z.RELATIONID_RANK, 03, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D3,
MAX(DECODE(Z.RELATIONID_RANK, 04, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D4,
MAX(DECODE(Z.RELATIONID_RANK, 05, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D5,
MAX(DECODE(Z.RELATIONID_RANK, 06, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D6,
MAX(DECODE(Z.RELATIONID_RANK, 07, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D7,
MAX(DECODE(Z.RELATIONID_RANK, 08, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D8,
MAX(DECODE(Z.RELATIONID_RANK, 09, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D9,
MAX(DECODE(Z.RELATIONID_RANK, 10, Z.FIRST_NAME || ' ' || Z.LAST_NAME)) AS D10
FROM
(SELECT FAMILY_ID, RELATION_ID, FIRST_NAME, LAST_NAME,
RANK() OVER ( PARTITION BY FAMILY_ID
ORDER BY FAMILY_ID
, RELATION_ID ) AS RELATIONID_RANK
FROM YOUR.TABLE
WHERE FAMILY_ID IN (123456, 654321)
ORDER BY FAMILY_ID, RELATION_ID ) Z
GROUP BY Z.FAMILY_ID;

The result will be in columns. The reason I needed columns was for an Insert into another table

Code:

123456 HOBO KELLY BOZO CLOWN YUCKO CLOWN SIDESHOW BOB
654321 ROY RODGERS DALE EVANS CISCO KID LUCAS McCAIN HOPALONG CASSIDY

I tried to line the data up so it would display correctly, but no joy.

Tried to edit for lineup, better but not pretty

Last edited by NASCAR9 on Wed Dec 19, 2018 10:48 am; edited 2 times in total

#2: Author: kolusu, Location: San Jose

Posted: Tue Dec 18, 2018 4:27 pm
—
NASCAR9,

You can use the trick of using XMLAGG listed in here

https://www.mvsforums.com/helpboards/viewtopic.php?t=12879

NASCAR9 wrote:

The reason I needed columns was for an Insert into another table

Isn't it simple to do a INSERT INTO with a SELECT statement? Something like this

Code:

INSERT INTO New_Table
SELECT FAMILY_ID
,CHAR(RTRIM(IFNULL(FIRST_NAME, ' ')) ||
CHAR(' ') ||
RTRIM(IFNULL(LAST_NAME, ' '))) AS FULLNAME
FROM Old_table
WHERE FAMILY_ID IN (123456, 654321)
;

#3: Author: NASCAR9, Location: California

Posted: Tue Dec 18, 2018 5:06 pm
—
kolusu,

I have xml to do the concatenating of data and flatting.
xml returns the data in a single column(as far as I know). I'm familiar with the link, I participated in it Very Happy

The point of my post was to show how to take vertical data and be able to retrieve the data into individual columns. This query is only an example I'm sharing.

Last edited by NASCAR9 on Tue Jan 29, 2019 1:37 pm; edited 1 time in total

#4: Author: kolusu, Location: San Jose

Posted: Tue Dec 18, 2018 5:36 pm
—

NASCAR9 wrote:

kolusu,
The point of my post was to show how to take horizonal data and be able to retrieve the data into individual columns. This query is only an example I'm sharing.

Nascar9,

I apologize for the oversight. Thanks for sharing the query.

#5: Author: NASCAR9, Location: California

Posted: Tue Dec 18, 2018 7:08 pm
—
kolusu,
No apology necessary. Your help is second to none.
I posted the query because I have never seen/used the Function DECODE before.
It solves a problem I'm faced with periodical. I'm hoping it's able to help someone else. There are many ways to skin a cat in SQL.

MVSFORUMS.com -> Database

output generated using printer-friendly topic mod. All times are GMT - 5 Hours

Page 1 of 1