Explain to me Dataflow Input/Transform/Output

Hi there, I am watching your video on Dataflows and I would like to know more about the process.


I see that you have input, transformaiton and output


I am confused because Transformaiton and Output are esentually the same. I could write an SQL statement that joins two tables together and gives an output. I am a little confused as to where I would put the join statement. 


Would I just put the join statement in the transformation and then in the output just type 

Select * from table; ? 


  • kshah008
    kshah008 Contributor

    Does anybody have any insight to @stevenmorey's question?

  • swyatt
    swyatt Contributor



    Yes you are correct about the "Select * " Statement.Your use of transformations over going directly to the output table is moreso about personal organization. While you can use joins in both segments I'd suggest doing the dirty work in transformation tables. Make sure you select 'Generate Output Table' at the bottom of the transformation dialog box. You won't select 'Generate Output Table' when you are altering tables or adding indicies. Those instances are where Transformations are most employed. 

  • kshah008
    kshah008 Contributor

    @stevenmorey, did swyatt's explanation answer your question?

  • Use the transforms to effectively create views of the data that you can use for following transform steps. 


    When you have the data in final transform, you can use the output to create a data source. What I've typically seen is that you will use your final transform step (this could be a single step, or at the end of multiple) as a view that you can select * from in your output table.


    This output becomes your data source, and is visible in the data source section of the data center.