Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi,
Today I discovered an odd thing in Qlik Sense.
I have a dataset where there are some rows that are exactly the same. All columns and data are the same for two rows. The dataset consists only of one table and no data modelling is done.
If I do a table with only dimensions (no aggregation), I only get one row of two. If I sum the data (amount paid), I get the sum of both rows.
I have checked the app in 3 different environments (prod, test and desktop) - all with the same result.
Can anyone of you explain this, and more importantly how can I fix this behaviour?
Thank you.
P.S. I can't share the dataset, as it is highly sensitive.
It's one of the main-features of Qlik that only distinct values are displayed in objects. It has nothing to do with hiding data else the data are intentionally stored in this way because Qlik follows no SQL-like row-based storing else it's a column-based storing and this leads to a significantly better performance.
If there is any reason to validate the data against the source-data (usually only during the development) it's useful to follow the suggestion from @mikaelsc to add recno() and/or rowno() to the loads - but after that they won't be needed respectively they become a disadvantage.
- Marcus
qlik will only show unique values.
add a differentiating field to your model and your table... you'll get what you want
(for example rowno() as UniqueRecID, )
Thank you @mikaelsc
Do you know if this is a feature in Qlik Sense? Can it be found in the documentation?
I am asking because I think it is odd to create this behaviour. I do not like when a programme "hides" data without me asking it to do so.
It's one of the main-features of Qlik that only distinct values are displayed in objects. It has nothing to do with hiding data else the data are intentionally stored in this way because Qlik follows no SQL-like row-based storing else it's a column-based storing and this leads to a significantly better performance.
If there is any reason to validate the data against the source-data (usually only during the development) it's useful to follow the suggestion from @mikaelsc to add recno() and/or rowno() to the loads - but after that they won't be needed respectively they become a disadvantage.
- Marcus