Do not input private or sensitive data. View Qlik Privacy & Cookie Policy.
Skip to main content

Announcements
Qlik GA: Multivariate Time Series in Qlik Predict: Get Details
cancel
Showing results for 
Search instead for 
Did you mean: 
Anonymous
Not applicable

Removing Synthetic keys

Hi

Could anyone suggest me a better way to remove synthetic key from the datamodel with following tables:

Available:
LOAD TESTER,
TESTER_KEY,
TESTER_CATEGORY,
TESTER_VENDOR,
LOCATION,
TESTER_VENDOR_AND_NAME,
TESTER_TYPE,
DAY_DATE,
DATE_KEY,
DAY_TYPE,
YEAR_QTR,
WEEK,
SUN_WEEK_START_DATE,
YEAR_MONTH,
YEAR,
HOUR_SEG_PRIME,
AVAILABLE_HOURS
FROM
[\\QLIKVIEW\PDS Data\QCT Analytics\Available.QVD]
(qvd);

Used:
LOAD USED_HOURS,
DATE_KEY,
DAY_DATE,
OPERATION_KEY,
OPERATION_NAME,
TESTER_CATEGORY,
TESTER_NAME,
LOCATION,
HOUR_SEG_PRIME,
SUN_WEEK_START_DATE,
TIME_KEY,
TESTER_KEY,
PROD_TYPE_NAME,
PROJECT_NAME,
SITE,
USER_ID,
JK_TESTER_KEY,
DEPARTMENT,
TESTER_VENDOR,
TESTER_VENDOR_AND_NAME,
DAY_TYPE
FROM
[\\QLIKVIEW\PDS Data\QCT Analytics\USED_HOURS.QVD]
(qvd);

1 Solution

Accepted Solutions
Oleg_Troyansky
Partner Ambassador/MVP
Partner Ambassador/MVP

This largely depends on the kind of analysis you need to perform with those two tables.

Basically, there are three ways:

1. Link Table - you build a table connecting those two, with all the common keys concatenated. This table will look a lot like your currebt Synthetic key table, only you generate the concatenated key explicitly.

2. Concatenate the two tables into a single "Fact table"

Those two methods are described in the standard "Developer II" training class, and also mentioned several times within this forum.

3. There is a relatively recent thinking that these kinds of synthetic keys are harmless, and are not so bad to live with. Personally, I'm of the opinion that we are still better off building our data models explicitly and not leaving the job of generating keys to QlikView, but the option of not worrying about it exists.

cheers,

Ask me about Qlik Sense Expert Class!

View solution in original post

3 Replies
Oleg_Troyansky
Partner Ambassador/MVP
Partner Ambassador/MVP

This largely depends on the kind of analysis you need to perform with those two tables.

Basically, there are three ways:

1. Link Table - you build a table connecting those two, with all the common keys concatenated. This table will look a lot like your currebt Synthetic key table, only you generate the concatenated key explicitly.

2. Concatenate the two tables into a single "Fact table"

Those two methods are described in the standard "Developer II" training class, and also mentioned several times within this forum.

3. There is a relatively recent thinking that these kinds of synthetic keys are harmless, and are not so bad to live with. Personally, I'm of the opinion that we are still better off building our data models explicitly and not leaving the job of generating keys to QlikView, but the option of not worrying about it exists.

cheers,

Ask me about Qlik Sense Expert Class!
mazacini
Creator III
Creator III

RE: Oleg's point #3 - you might find this post interesting.

http://community.qlik.com/forums/p/25440/97137.aspx

Joe

Anonymous
Not applicable
Author

Thank you both for the responses , i guess i can live with the synthetic keys as i dont see any performance issues and data issues as of now with my data model..