I have been working to develop a treemap visualization in Qlik Sense and have hit a bit of a snag.
I've created a drill-down dimension, consisting of (Function, Category, Subcategory, and Findings), that the chart is viewed over. The corresponding measure is Count([Findings]).
Findings are either good (positive) or bad (negative) and should be colored on a two-tone scale (I chose red and green) according to the value in a corresponding field (Impact, equal to either 0 or 1).
At one level up - at the subcategory view - and onwards, I would like the coloring to change to a 4-toned scale (1-4) which corresponds to a rating, another field.
Findings are not associated with a rating. There is no set pattern for which combinations of good and bad findings correspond to a particular rating.
Initially, I coded for the findings to be colored first, followed by the other dimensions in the drill-down. This simple solution worked in most cases, but, for example, due to the method (unknown to me) by which groupings inherit color, a "Sub Category" with all "Positive" findings was inheriting the green color of its children findings, overriding the blue "3" rating.
Now, I have created an expression that almost works: in the above cases (where only one 'type' of finding is associated with a particular subcategory), only the finding that appears first in the raw data is inheriting the color of the parent 'subcategory' while every other one is colored correctly.
I realize this is a some what long-winded request for what is ultimately a review of my logic / understanding of the BNF used in Qlik expressions, so thanks to the community in advance. I have attached my current expression to this chat.
