Hi Bob,
A very tough requirement, since the dataset is not well organized for pivot purposes. However you can try these:
1. Create one category with grouping expression: =true - e.g. make a single data cell for the entire table.
2. For each choice that may appear in the list (Vanilla, Strawberry etc.) create a Values data groping with this expression:
=SUM(TONUM(LIKE(SPLIT(Fields!Produce, ";"), "Vanilla")) * TONUM(SPLIT(Fields!Quantity, ";")))
Lets explain this expresison a little. The split fuction splits an array of strings to another array of strings by a given delimiter.
The Fields!Produce and Fields!Quantity in our case return arrays, containing all records from the data set. If we have this set:
Vanilla; Bananas - 10; 20
Oranges; Vanilla - 30; 10
These functions will return these arrays:
Vanilla, Bananas, Oranges, Vanilla
10, 20, 30, 10
The LIKE function is a chain function, so we transform the first string array into a boolean array the arrays become:
true, false, false, true,
10, 20, 30, 10
The multiplication operator is also chained - so we multiply the two arrays giving us the result array of:
10, 0, 0, 10
Finally we sum with SUM funciton. So for the bar Vanilla we get a value of 20.
The other Values Data Grouping (for Bananas, Oranges etc.) simply have the same value expression but with Vanilla replaced with Bananas, Oranges etc.
So you can create a bar for each type of product and they will appear in a single cluster. Unfortunately you can not create a separate category, but you can use the legend to show the series - which will do the job we think.
Best Regards,
Nevron Support Team