Filtering for a subset of data from a single table

[mlauber71]

@mvaldes I tried a few things and came up with this. But I fear if your Excel files do not al have this exact structure it can get more complicated. You might try to better identify the headers of the column if they might shift. Also the two screenshots you provided do not seem to be from the same file.

So I think you will have to invest some more work.

Key to the approach might be to mark the blocks you have by a rule filter (what would a start of a table block look like). In this case it is the term “DAY*” in a column

then copy the RowIndex to all the cells that are following so you can ‘mark’ a table block. And you can take it from there. There are some more complex examples on the forum. And since there questions appear a lot lately I might compile a list or a tutorial. You might search the forum until then.

try to untagle a complex Excel file – mlauber71

Maybe take a look at this example also:

How to change human readable catalog to basic spreadsheet format

Hi @DWJames , there are a few ways to do this, and can be done without loop. But without a loop, it will become tedious to clean up some rows, one column at a time if I process them all at once (Knime process data row-wise, meaning if I have to clean up 7 columns, I need 7 operations, potentially 7 repeating nodes, or eventually use a loop). I’m actually going out of my way to include the empty rows between the brands This is what I came up with: [image] Input data: [image] Outpu…