I still think that there's a way to use ExecuteSQL to produce the data you want so that you can do a chart from delimited data, but I've been playing with different queries against your sample data off and on for over a day and still haven't come up with one that works.
An alternative approach would be to set up a table where the duplicate do not exist and thus do not need to be omitted.
And I suspect that the "sum the reciprocal" method could be adapted to the needs of this chart: How to count the number of unique occurences in field.
But an SQL query still "feels" like the better option if someone can just come up with the right query that will work...
Thanks for the info. I actually tried something thatI thought would work, essentially marking duplicates then working off that result. Ill let you know if it works :)
Yes, if you perform a find to omit the marked records, that should work.
It did work - using the Mark method! Thanks!