Skip to main content
Announcements
Qlik Connect 2024! Seize endless possibilities! LEARN MORE
cancel
Showing results for 
Search instead for 
Did you mean: 
curiousfellow
Specialist
Specialist

Sorting in a chart

In my chart the first column contains text values like this '100_110', '100_120'....My chart has to be sorted by the value of this column

I made this column conditional visible. When the column is visible sortings are correct, when it is not visible one value is sorted wrong.

The chart displays the values of one table. This table contains values from multiple excel files.

All excel sheets have the same structure, loading is done by load from   .../*.xlsx

The row that sorts wrong is from a file only containing values for that row, the other excel-files contain multiple row-values.

I tried to sort the chart by text_value and by expression too, but with same result. 

We are using Qlikview May 2021 SR 1 Version 12.60.20.100.0

 

 

Labels (1)
3 Replies
rubenmarin

Hi, can you upload a sample app with some dummy data that demostrates the issue? That value should have something different, and it will be hard to guess without the sample to check.

curiousfellow
Specialist
Specialist
Author

An example is hard to create. I don't have the time for that  at this moment (deadline) 
I was searching for differences too, but cannot find any. And should it not be sorted wrong too when it is visible then ??

I gave the first column in the second file the same value as one of the records in the first excel file. Sorting is still wrong when this column is hidden. I copied the chart to excel as full table and compared the values. Excel cannot find any differences.
The reasons I can think of why the sorting is wrong:

1. It has something to do with the load order.
2. The first time I loaded the files the formatting of the fields in the excel files differed : first file the values contained an underscore the other files contained minus sign

 

 

marcus_sommer

That's in general the intended behaviour because to exclude hidden sheets/objects/objects-parts from any calculation is saving resources - which seems also to happens in your scenario.

But there are different ways to "hide" things which have also differences in the behaviour, for example by applying a condition to a dimension vs. to hide the dimension per tab presentation which a straight-table supports but not a pivot.

Beside this you may transfer the sort-order per expression to the available dimensions or preparing an appropriate sorting per load-order in the script or with any dual() logic to the dimensions.

Two screenshots of the chart with/without the mentioned dimension would be useful to understand the wanted goal.