Pivot table: expressions as rows, dimensions as columns, how to do it without trial and error?
Dear fellow qlikviewers,
I have created a pivot table (chart object), with some dimensions and expressions. For the sake of formatting, it would be much nicer if the table is rotated, since this makes the used vertical space constant. I have occasionally succeeded in dragging the dimension rows to the top, to make them columns, and the expression-columns to the left, which makes them rows. However, just as frequently, the table will jump back to its original orientation (after hiding a dimension or adding some expression for example), after which it seems to take a random number of these dragging operations until order has been restored in my qlikview universe. I'm at the point that I believe the precise pixel and amount of time for the dragging operation are responsible, and before I extend an exhaustive search on those, I thought I would ask around if anyone here knows how to accomplish this consistently... You would help out a fellow qlikviewer enormously
To the best of my understanding, once you've pivoted some stuff, QlikView will keep things pivoted depending on the dimension/expression number. Meaning, if you've pivoted the 2nd dimension of your table, and then the 2nd dimension is either deleted or conditionally hidden, then if there is a 3rd dimension (which has now become the 2nd dimension) it will get pivoted, and if there is no 3rd dimension then no pivoting will occur at all. The same thing would apply to expressions, I believe. The key to maintaining order would then be to ensure that the dimension/expression you want pivoted is always in the same position no matter what is hidden.
Apologies for my late response, I was off for a long weekend, but thanks for your response! I was doing a sort of drilldown with the pivot table (there is some branching going on, which is why I can't do an actual drilldown), but based on your response I rotated based on the first dimensions and expressions. This worked after the third attempt and drilling down seems stable enough again. That is much faster than before, although I recall also having made succesful attempts starting from the bottom dimension. Nonetheless, it's working again, so thanks for your input!