I´m new in the “world” of KNIME so please be indulgent with me.
I have a table like this:
Column Header: A B C D E
1 Product 1 True Product 2
2 Product 1 Product 2 True
3 Product 1 True Product 2 True
And I want the table like this:
Column Header: A B C D E
1 Product 1 True
2 Product 2 True
3 Product 1 True Product 2 True
I tried the Rule based Row Filter and filtered the rows for the “true”, but then nothing changed because every row contains at least one “True”.
So I thinking about using different Row filters for every column and mix the tables in the end again. But is there a smarter and faster solution for this?
I hope i’m understand your data correct.
You try to remove entries where either in column C or in column E are empty (not True) and shift column D and E to column C and D if C and D are empty?
Further do you only have this 4 columns to check or does your real dataset contains a lot more columns to check?
yes you understood it correct. I´m try to remove the empty entries (not True) and shift the other column to the front so i have no empty entries in between.
I have in total 6 columns to check.
Based on your sample data, I can’t even tell what makes a column.
Is “Product 1” one column, or actually 2 columns “Product” and “1”? In the event that a space is part of the data, can you instead use a delimiter other than space? In any case, it is perhaps better to use a delimiter other than space in any case, as it is hard to see the empty cells with space as delimiter.
hopefully this example helps you to go more inside. I’ve build it on a generic approach, which is independent of column numbers. In addition you could modify the math formula to bundle e.g. 3 columns instead of 2 columns by changing the formula from floor(($COLUMNGROUP$+1)/2) to floor(($COLUMNGROUP$+2)/3).
