Hi everyone, I would like to filter the rows as the rules below: -If “code” column is starting with “S1”, “stock” column is “0” and “order quantity” is greater than “0”, bring all the “Group” rows of that group. -If not, eliminate all the codes of that group. [Exp.] For example in the 3rd row, “code” is starting with “S1”, “stock” is “0” and “order quantity” is greater than “0”. In this case I want to bring all the rows of “A” group and I want to eliminate the B and C group rows. I want it to be like the table below; [exp2] Thanks !

Hi @ellobyhn , one approach is to create a rule based row filter that finds all rows that match your rule. So this finds the rows containing Code beginning S1, Stock=0, Order Quantity > 0 You can then join this back to your original data set on “Group” and that will bring through all rows for the required group and no others Here I’ve also used a groupby on “Group” as a shortcut for ensuring I don’t have multiple rows containing the same “Group” coming though. [image] These “required groups” can then be joined back onto the original data, by “group” so it brings through all required rows, but none of the others There are other ways that this could be achieved but hopefully this gives you an idea to work with. filter using rule and join.knwf (11.1 KB)

You can use [image] Nominal Value Row Filter — NodePit Filters rows on nominal attribute value or [image] Rule-based Row Filter — NodePit Filters the input table based on user-defined business rules

Hello izaychik63, Thanks for your reply. I have tried the rule-based row filter but I want to filter these table on group based. For example even if the second row doesnt meet the rules I mentioned, this row should’nt be eliminated as they are in the same “group” (A group) with the third row.

Hello takbb, It works! Thanks! :slight_smile:

KNIME Forum Archive

Filtering the rows by rules

KNIME Analytics Platform

izaychik63 May 23, 2021, 6:21pm 2

You can use