Hi everyone,
I realise this question has been asked before, and I've tried several solutions from here, but somehow I can't get them to work for me. My only wish is to be able to let the user map the columns of their data coming in to our data model so we can then dynamically analyse everything correctly, even if the columns originally had different names.
So user workflow:
This is sort of what my workflow looks like right now:
I attached the example; ideally I would have 5 options for 5 columns that need to be mapped/renamed.
I'd be super grateful for any help!
Cheers,
M
@Inactive User , in that case you can adjust the version in the workflow following the instructions in the following link:
Outside that, we can't help you until you upgrade, I'm afraid.
M.
I managed to open if now, however, it doesn't seem to work:
Now every column has the wrong name. 😄
Is this because it's renaming by position and not name? First question should rename Sum to Profit, second Report Date to Date etc.
Thank you for your help and patience.
@mceleavey it seems that the column order gets messed up during the cross-tab.
This is how they go in:
after the transpose:
(order still correct)
and how they come out of the cross tab:
I'm guessing the order out of the cross tab is just alphabetical (after User ID)? The union then messes them up because the positions are different. So it seems that this won't work, since I can't find an option to un-order the columns?
@mceleavey It works! Thank you so so much!
@mceleavey Oh no wait, I just realized that it all depends on having one column that's distinct and that we know the name of (for the first transpose). Right?
Is there a way to do it without knowing any names of the incoming columns?
@mceleavey Sorry if that was unclear, I mean that in the first activity, the transpose tool, you transpose it using an existing column. If the input is a different file with a different schema and there is no column called 'User ID', this would not work. Right?