1 of 1 people found this helpful
There is a possibility that the method described here may
suit your purposes:
Basically, it makes a string of all the possible 2nd dimension values
by making a running list using PREVIOUS_VALUE() + Current value.
If that doesn't fit your scenario, would be grateful if you would post a
sample workbook with just a small amount of fake data, in a format
similar to your true set.
Thank you for the sample workbook.
IF (PREVIOUS_VALUE("")) = "" THEN ATTR([Color])
PREVIOUS_VALUE("") + ", " + ATTR([Color])
And then in a separate field made
a WINDOW_MAX per Person.
and it seemed to produce the needed result.
I also added a few rows of data to show that
it groups the persons by motif regardless of the
the order of the colors in the dataset.
Please see workbook v2018.1 attached in the Forum Thread.
I edited the above post.
I think the Previous Value calc
and the Window_Max calc
needed to be in separate fields.
This method did work for me. However, I ended up going a different route given my back end use case. Since I connect to an SQL instance, I wrote a new custom SQL query as such:
person, array_agg(distinct color) as palette from mytable
group by person
All the best.