Pivot table expand/collapse not working consistently

[#1]

I have a pivot table with 4 dimensions and 1 expression, pivoting is disabled.

The right-mouse click features Expand All/Collapse All work fine. But when I go through the table, expanding and collapsing individual cells, I run across some cells that have erratic behavior. For example, I see expanded cells that refuse to collapse (other cells in the same dimension will collapse properly) and I see collapsed cells that refuse to expand (other cells in the same dimension expand properly). If I use Expand All, I see that they all have sub-dimension values; i.e., they should be individually expandable/collapsable. Oh, and their [+]/[-] buttons do change state, its just the cells themselves that don't.

The same thing happens whether I'm in the desktop or access point.

Has anyone else seen this? If so, any idea what the cause is and how to fix it?

Thanks.

[#2]

Just to add a little more info, when I open the report, some of the collapsed cells have the [-] icon and some of the expanded ones have the [+] icon and these are the cells that turn out to be problematic. Obviously the cell's initial state as displayed by the icon, doesn't match their actual expanded/collapsed state. And no amount of clicking on them gets them back to behaving normally.

[#3]

i have the same issue constantly and since i am evaluating the product as of now. I am afraid to show it to users because of these bugs

[anthony_kinsell]

Hi,

Did you find a solution to your problem?

I had something very similar (I think)

See link to my query below:

http://community.qlik.com/message/412686#412686

[#5]

what i noticed is that there seems to be a trick to it. It seems that if there is at least 1 expanded field ANYWHERE in the chart, it will have those issues. So to make sure that i dont get that issue, i have to collapse ALL for each field till i get to the last one then it works properly again.

[#6 · Accepted solution]

Turns out the problem was caused by the fact that we had some null values in our data. Once we adjusted our model to get rid of the nulls, the problem went away.