Qlik Community

QlikView Deployment

Discussion Board for collaboration related to QlikView Deployment.

Not applicable

Performance: Filtering at Scripting Vs Filtering at Dashboard

Hi All,

Am doing the above work and am at a dilemma.  Do advise, below is the scenario:

I have 2 set of evaluations data in my Qlikview dashboard.  Both of them have a column called timestamp. 

Questions:

1.  Should I create individual column of timestamp for each of this evaluation?  Example.  Column A called A_TimeStamp and Column B called B_TimeStamp in one concatenated table.

So at dashboard level, i can just create a listbox with A_timestamp for report A and B_timestamp for report B without any data filtering.

OR

2.  Should I concatenate them as one column called timestamp. 

So at dashboard level, I have to create a listbox with expression to filter out based on the unique identifier of the evaluation.

Which is better in term of performance ?

Please kindly advise as the data will be getting more and more.

Thanks in advance.

5 Replies

Re: Performance: Filtering at Scripting Vs Filtering at Dashboard

The only way to make sure which option performs best for you is to test both options. My guess is that it's better to concatenate the data into one table if both sets contain the same fields and add an addition field to the table that indicates where the data came from:

Evaluations:

LOAD *, 'Evaluation A' as Source

FROM ...evaluation_a_source_table;

Concatenate(Evaluations)

LOAD *, 'Evaluation B' as Source

FROM ...evaluation_b_source_table;


talk is cheap, supply exceeds demand
Not applicable

Re: Performance: Filtering at Scripting Vs Filtering at Dashboard

Hi Gysbert,

Apologies for late response.

So given your suggestion, if I concatenate these 2 tables into 1 table and create an additional flag to indicate the difference.  Then in my dashboard, I need to use the IF expression in order to segregate out the difference ?

If yes, is there any other functions other than IF expression that I can use in this context?  The reason behind asking this question is that IF expression is costly too many if statements are in my dashboard.  This is why I tried to minimize this IF expression usage.

Your advise is greatly appreciate.

Thanks

Chun Wey

Gabriel
Valued Contributor II

Re: Performance: Filtering at Scripting Vs Filtering at Dashboard

Hi,

Yes IF function is costly, however what you can do is SET ANALYSIS for filtering in your dashboard.

Gabriel
Valued Contributor II

Re: Performance: Filtering at Scripting Vs Filtering at Dashboard

I just realised that you've asked same question in another post which you've marked as ANSWERED.

Not applicable

Re: Performance: Filtering at Scripting Vs Filtering at Dashboard

Hi Gabriel,

If you are referring to this:  IF Statement Performance at dashboard level

They are similar (both in term of performance) but the questions are asked differently.

For the link, it is asking about the IF performance at dashboard level and how to optimize it.

In this thread, it is asking about the approach to data modeling to prevent the IF statement. 

I hope the above clarifies.

Thanks

Chun Wey

Community Browser