ApplyMap load speed vs lookup table

[#1]

Good day all

I have a large data set that contains many acronyms and abbreviations. I would like to convert these into their long text translations.

My question is: is it faster to ApplyMap these values directly into the FACT table or keep the values in their own lookup table?

[manojkulkarni]

ApplyMap is much faster than LookUp Function.

Go thru this link for detailed information

Joins and Lookups

[tresesco · Accepted solution]

Keeping them in their own lookup table would obviously be faster, because that doesn't require any comparison, it is just a simple table load. The comparison could have been a valid one had you thought of joining it or just mapping it, where the apply map would have been faster.

If you keep them in separate tables, the load would be faster, however, the front-end performance would be a bit slower than it could be they were joined or apply mapped to form a single table.

[Peter_Cammaert]

That all depends on the size of your data. The difference between using a translated field in the same table as the original text vs a separate normalised table through a link is minimal, according to the Qlik gods. See A Myth about the Number of Hops

[rustyfishbones]

ApplyMap is faster, use Lookup if you need to add more than 1 field to your fact table, although you would better off having a number of Mapping Tables and using ApplyMap

[#6]

Thanks for your reply. So, leaving the long text as an associated value in a seperate table would be quicker in the front end than using mapping tables?

[#7]

Sorry, I should clarify; I wanted a comparison with a lookup table, not the LookUp() function. Thanks for your reply.

[tresesco]

No, it's otherwise. Faster in the back-end. And slower in the front-end, but that could be negligible based on your data size.