Qlik Community

Ask a Question

New to Qlik Sense

If you’re new to Qlik Sense, start with this Discussion Board and get up-to-speed quickly.

Announcements
QlikWorld May 10-12, Online and Free! REGISTER NOW
cancel
Showing results for 
Search instead for 
Did you mean: 
Gopi_E
Creator II
Creator II

How to Create Master Calendar and Data Model when we have 2 dates coming form 2 different tables

Hi Qlikers,

I have a data like below

Table1(available only last 12months data respected dates)
Id,
eventData,
ProcessDate,
comments,
Flag1

Table2(more than 12months data)
ID,
eventDate,
ProcessDate,
ratings,
Flag2,
....,
....;

In above tables associated by ID and created flags in each table to identify the table and number of months. when we create master calendar using above  2 common dates which come from table 1 and table 2, the Table 1 showing more than 12 months data.

1. the first problem is when we create master calendar using Table 2 which has more than 12months , when i click on Table1 flag, the months were showing more than 12months data(Table 1 should has only 12months respected dates)

@sunny_talwar - can you suggest any idea 

-Could anybody help me out how we can create master calendar using above two dates which come from two different tables and which type of data model would give correct results

Thanks in Advance

 

4 Replies
marcus_sommer
MVP & Luminary
MVP & Luminary

The easiest way is usually just to concatenate the fact-tables and not to associate them with any linking - just give it a try.

- Marcus

Gopi_E
Creator II
Creator II
Author

Hi Marcus,

Thanks for the quick response

I have tried that technique, Its creating extra more number of millions records with null values, so the loading time is increasing.

 

Thanks,

marcus_sommer
MVP & Luminary
MVP & Luminary

That such table is in some degree asynchron which results in NULL for some fields in many records is often not a problem because these NULL's won't be stored - it only means that there are no values available.

Without any record-id's which are only very seldom useful within a final datamodel (during in evaluation/design process they may helpful to find any differences) the max. pointer for the data-table shouldn't significantly increase even if there are much more records included and therefore the run-times shouldn't be much longer especially not if not only the loadings are compared else also the efforts to calculate the datamodel.

Quite often there are also real benefits while the loading because no complex key is needed (you have at least two fields which needs to be combined) and you could apply simple dimension-tables and not any complex constructs mostly done with link-tables between the fact-tables on which the dimensions are linked.

Therefore I suggest not to discard the simplest solution too fast. I use this concept very often and not only because of it's easiness else because of the good performance in script and UI.

- Marcus   

Gopi_E
Creator II
Creator II
Author

Thank you Marcus !!

This will help me to understand the data insights.