Where is better to create new fields

[Anonymous]

Hi All,

my qlik sense application is very slow, maybe this happen because it is full of measures and dimensions created in the master item and not by script. My question is the following:

"Is it possible that is more powerful to create complex dimensions and measures as field directly in the data load editor instead to create them in the master item? By creating all of them in the script the calculation time could be reduced?"

Thanks,

Giulio

[MK9885]

I'm not sure if this would help

1. Extract the raw data from your data source and store them as qvd's

2. Do the transformation by loading extracted qvd and store them as transform qvd's

3. Load all your tables into one QVF file, which will be your Data Model file.

4. Use your Data Model qvf file as Binary Load and do your designing in source file.

5. Store big/complex expressions in variables.

QVD's load faster and decrease the size of the file and it's better if they are optimized qvds.

If it is still taking time to load then check with De bug mode available in your script to know where exactly it is taking time? It can also be due to number of rows you have in your data set.

[Anonymous]

My problem is the calculation time in the "Presentation Layer" I see so many times tables looping and I am just ask if I can reduce this loading time by pre-calculate the fields in the Data load editor. For example my app is full of aggr( function.

[MK9885]

Aggr cannot be done in edit script.
You've to do it front end. Try using any other alternate function to calculate your measures with same result?

Or store your measures as variables and try it

[Anonymous]

but aggr can be replaced by a group by statesmen in the data load editor, I just want to know where is better to do the aggregation in terms of performance in the front end or directly in the script?.