Link Table in Qlik Sense

Link Table and a use case
Outline
- Customer had multiple date fields
- One primary key is related to multiple date fields
- How does one do analysis with multiple date fields?
o This requires some exposure to a Master Calendar
o Relate the Master Calendar to a Link Table - a table that incorporates multiple calendar date fields
o Apply Qlik’s powerful set analysis capability
I was recently working with a prospect in the quality assurance (QA) space. The customer had an interesting problem on his hands. Whenever a QA event arises in his team, a QA number is generated. However, each QA #, at some point, would have a Date Identified, Date Initiated, Date of Occurrence, and Date Closed field. He was stumped then on how to perform analysis on his data with these multiple date fields. For example, if he wanted to know how many events were Identified, Occurred, and Closed, in August 2020, how could he do this. He knew he would need some arbitrary master calendar that connects all these date fields together. But, that’s not all, he still needed a way to relate all the four date field types to the master calendar.
We resolved his problem using Link Tables and the video below goes through how this can be implemented. In this example, I am playing the role of a retailer, think of Amazon, whom for any order generated with some Order ID number, there would be multiple dates related to this order. The dates here are OrderDate, InvoiceDate, ShippedDate, and ArrrivalDate. Watch how easy Qlik makes analyzing this data set easy. Additional supporting resources can be found in the companion blog where this video was posted.

[@abdullah01015]

Concise! Thank you very much!

[@alexis1959]

Great solution but as someone else mentioned LOAD DISTINCT would remove the duplicates that we see in the Link table..

[@itziklusky2007]

Thank you it was very helpful!

[@mukulsharma8568]

Thanks it's a good one

[@ChrislyBear]

Would it make sense to "Load Distinct" from the orders table? It seems to me, that the link table may contain duplicates, that aren't needed for the connection, right?