Data model issue - inconsistency - how bad is that?

[datanibbler]

Hi,

my issue is the following:

- This year (2014), some dpts have introduced new Excel_lists, among them the one on (the app of) which I'm currently working:
- I am and have been loading the data into QlikView as many tables although it all comes from just one Excel file, just to keep it clear.

  => In most cases, I could append the list from the 2014 file to the list from the 2013 file since the structure is the same.

<=> In the current instance, I can't do that because there are many more fields in the list for this than in the list for last year.

=> My question is this: Can I concatenate them still (forced concatenation)? That will generate a lot of NULL values I know.
    <=> On the other hand, it will make the whole thing a lot easier because I won't have to implement two redundant expressions
    - using different fields (from different tables) for 2013 than for 2014.

Can someone clarify this for me? What exactly are the consequences of this structural assymetry?

Thanks a lot!

Best regards,

DataNibbler

[marcus_sommer · Accepted solution]

I think forced concatenation is the best way in this case - NULL didn't matter. I do it in a lot of cases.

- Marcus

[datanibbler]

Hi Marcus,

yes, I am doing that and so far it seems to work. I just had to flexibilize a few of the expressions on the GUI to work based on a different set of fields, depending on the year the user has selected.

Let's see.

Thanks a lot!

Best regards,

DataNibbler

[marcus_sommer]

See also: Performance w/ multiple fact tables - star sche... | QlikCommunity