You can use Tableau to quickly find mistakes --
as you've done already using your second view.
But to repair mistakes, the new (correcting) transactions
should be added to (written into) the original datasource(s).
Tableau can't do that (it is not even indented to do),
but your 'transactional application' can --
with the help of a human, of course ;-)
Thank you for your response.
Due to business processes, I am unable to write back changes to the datasource so it appears that in this particular instance I may be stuck.
You have confirmed by fear, but it does mean my thoughts were right about Tableau and at least I can say I am learning something! Thanks again.
Rather than blending, could you join the Invoice and Credit tables with an outer join? That way you would tie the invoices and corresponding credits together, but would also get all invoices without a credit, as well as all credits without an invoice. On the surface, it appears that your tables are in the same database.
Thanks for your response.
I too thought a join would work but the data sources are collated differently and I get a collation error message when trying to join. I don't know SQL so I don't have a way around that issue either. Hence the blending.
That's strange, most of the issues I can find with collation errors happen with Excel or csv files rather than databases. But yeah, with the way blending works, I don't think you'll be able to get the result you want with a blend. Could you possibly reach out to your IT team to assist in a quick outer join query that you could pull in with Custom SQL?
Sorry can't be of more help!