How can I skip null records

Question

Hi all,

I've got a strange question here. We're comparing vendor data to internal data, and for reasons out of our control we have to convert a scanned PDF to an excel table (using able2extract). The problem is that any mark or crease on the scanned PDF can cause weird data to appear. I'm cleaning up the data in alteryx, but I'm trying to basically "move all data left" in each row, ignoring null/blank cells.

A visual example:

The left side is how the data looks now, and the right is how I want it to be.

I found This thread here but that moves all data up, not left.

The excel equivalent of doing this would be to select blank cells, then delete selection shift:=toleft

Can someone help get me started with something like this? I'm trying to avoid cleaning it up in excel first, and I think I can clean it up adequately after this.

Thanks

mvidas · Answer

The field names don't matter and I'll be changing the format anyways. I imagine the right-most columns will all be empty but I don't need them there if you have a way of also getting rid of them.

The PDF->excel conversion makes the excel file unpredictable, and column count can vary from page to page. The scanned invoice really only has 10 columns on it, but each page converts differently; some can have 10, some can have 15. It's really an awful process. I'm hoping by compressing each row "left" will let me find some consistency.

JagdeeshN · Answer

@mvidas

The current state(Left side), does that retain the same state(with empty columns) and format throughout and do the field names remain the same too?