Run an Union in dataflow with two datasets with different columns

Persio DeLuca-Oracle

Summary:

I want to run an Union operation in dataflow with two datasets with different groups of columns and some columns (date and filtering dimensions) in common.

Content (required):

I am getting an error message saying that Union is only possible with two exact tables are needed, including the order of the columns in the dataset, which sounds weird since any other tool can execute that easily. Am I doing something wrong? Is there another way to achieve the same results? I don't want to merge since two lines do not represent the same event, but possibly two different events in the same time (the events are not related, but I wanted to filter by a common set of dimensions).

Version (include the version you are using, if applicable):

Code Snippet (add any code snippets that support your topic, if applicable):

Bret Grinslade - Oracle Analytics-Oracle

The Union functionality currently has this limit. In your situation --you may want to try creating one datasets with multiple tables or use blending. In data flows today - the Union operator is trying to create a fully denormalized data table in the flow. The other option you have is to add and select columns to each side so they have the same columns, types, and order -- but this might not be what you want.