OK just to see if I am going down the right route...
To handle possibility of duplicity and to avoid need for CSE Array I added the following:
This way you have unique numbers for all - but the variance between BN & P is minute and can be removed using Rounding whilst at the same time allowing you to differentiate between states with same value in P.
Which for me generated the following output:
Using the LOOKUP approach I hid error returns where there were not sufficient number of values to return - eg BL7 there was only one prior year value > current year value - namely North Dakota so only 1 value returned in the string.
If the above is along the right lines we can adjust for the Lower values on a similar theme.
Bookmarks