Your data from Excel has imported as numbers and it's representing itself incorrectly. I.e. 14591920383T
You need to change the datatype's on the fields in Excel to textor in the tableau data connection dialog, change them to string. In the Tableau Connection dialog on the header of the column you should see a hash/pound icon next to the column name. Click it and change it to string.
Take a look in the screenshot you've attached; '1.46215e+10'. Surely this isn't correct?
I tried to attach the new file here but did not find the way.
Anyway, I added a "-" character to each transaction ID in order to fix that. But it doesn't change the problem.
My problem is that when I create a calculated field in the Transaction Code source and try to call the cost in the other source, Tableau obligates me to use ATTR. I don't want to use aggregates here, but exact number for each transaction code whenever there's a match.
Do you know how to solve it?
When you reply, there is an option in the top right hand corner of the reply box "Advanced Editor and add an attachment".
If you're blending like this you need to perform some sort of an aggregation. ATTR or MIN usually gets the required result. What's wrong with the results returned by ATTR?
I've reattached the docs, including a desired outcome tab in the excel sheet.
My real case is much more complex than this example. I need to execute several steps of calculation and some steps require information from other data sources. The calculation needs to be executed on "row level" in the primary data source. So each row will contain an unique information, which is not possible if I bring an aggregate information from another data source.
Do you have the solution to solve the problem in the example I've attached?
What is your problem?
I just looked at your transaction code sheet and took the ID's for the first three records which were not #N/A. I looked them up in Tableau, they match exactly to your worksheet.
It's not clear what you're problem is or what you're trying to solve.