Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi all,
The structure of our data is something like
Account can have a salesperson
Invoice can have a salesperson
Salesperson has an id which is linked via account and invoice, and that id has a name next to it, what I would rather not do is do an apply map as there is additional data as part of this table and I believe applymap only just vlookups the id against a descriptor.
Otherwise it creates then syns, how would you recommend I bypass this problem? As syns I presume are not a great thing to have? As they seem to be junction tables to turn a many to many into a 1 to many relationship. Otherwise I am just creating a brand new table with the same data but a different name to get this working and it feels inefficient.
Many thanks,
James.
It looks like you have 2 roles that a person can perform :
For your visualisation requirements is there any need to have person as a dimension and have expressions relevant to both roles ?
I would suspect not as this would probably not return logically results.
An invoice needs to be there for historical significance as an account can be updated. I guess the only way is to duplicate the tables?
Your SalesPerson table only has 3 fields :
Z_ID
Name
Group
So you could do a couple of Mapping Loads
Using Z_ID to return Name
Using Z_ID to return Group
Then a couple applymaps whilst loading the Invoice table creating 2 fields
Invoice Sales Person
Invoice Sales Group
And a couple more applymaps for the Account table creating
Account Manager
Account Manager Group
Thus you'll have no SalesPerson table in the resultant data model. ergo no synthetic keys or loops.