Dealing with nulls in TI
Posted: Fri Jul 06, 2012 10:36 am
Hi there,
Is there a way of dealing with null element values in a CSV file when importing in TI?
I have a CSV file with a column that is not always populated as it's not always relevant. The TI process runs but does not import the rows that have a null element - error is : Data Source line (21) Error: Data procedure line (4): Invalid key: Dimension Name: "SubNominal", Element Name (Key): ""
I would like to replace each null with something like "n/a" and have this as an element in the destination dimension, and I want to avoid having to do this in SQL before TI gets the file as this is to be a direct export from the source system.
Can it be done in the prolog section or metadata section with a rule?
I am quite new to TI code - I've had a good trawl round the forum and the TI user guide but I can't find anything specific that doesn't recommend doing it outside of TI.
Thanks
Steve
Is there a way of dealing with null element values in a CSV file when importing in TI?
I have a CSV file with a column that is not always populated as it's not always relevant. The TI process runs but does not import the rows that have a null element - error is : Data Source line (21) Error: Data procedure line (4): Invalid key: Dimension Name: "SubNominal", Element Name (Key): ""
I would like to replace each null with something like "n/a" and have this as an element in the destination dimension, and I want to avoid having to do this in SQL before TI gets the file as this is to be a direct export from the source system.
Can it be done in the prolog section or metadata section with a rule?
I am quite new to TI code - I've had a good trawl round the forum and the TI user guide but I can't find anything specific that doesn't recommend doing it outside of TI.
Thanks
Steve