INNER JOIN vs WHERE clause

[Vladislav_qv_dev]

Hi, everyone!

I've got a question. I have a table with large number of rows and I want to reduce it by 1 single value of 1 field, i.e. I want to have in a new table only rows with value "Pepsi" in field "Brand". What solution is the best for performance: to use WHERE clause, while loading data in a new table OR to create a new a table by inline load with 1 field and 1 value ("Pepsi") and use inner join between them?

Thank you in advance!

[kavicsqss]

Hi, I am currently optimizing a qvw app with approx 4,1 M records. It is big enough to see the performance impact (reload time) if script is written differently.  I have made a lot of time-reduction by creating an intermediate qvd table earlier, then loading the table from QVD optimized (basically without any change), then loading 1 field with 1 value (like you said, or with autogenerate), and inner joining this small table to the big one. (instead of Where)

If Where is more complex, e.g. Where Field = 'A', Field2 <>'B', etc., I first filter the big table with inner joining Field (value A), and then do a 2nd load with a reduced number of rows already and then apply Where Field2<>'B'. This of course depends on how well inner joining Field (value A) reduces the size of the big table.

I 'd be pleased to hear from fellow Qlik-ers to learn from their experience as well 🙂