For col1= 18 and col17=133 several rules may fire. Not interested in only firing 1st one that matches.
There are many such rules. I need to fire all rules and then to ascertain how many times the consequent occurs. This will be the most probably value for my input dataset. It's form of prediction/classification.
You can put your rules to a table, iterate through them with a tablerow to variable loop and in the loop you can set the rule based on the flow variable (you have to have at least a comment rule to have an array in the flow variables tab). Based on those, you will have the outputs as missing or the values after the loop. The non-missings should be replaced with 1, the missings with 0 and after that a group by a (set of) key columns will give you how many rules fired for each row. Would this help to solve your problem?
Cheers, gabor
PS.: Or just create a new node based on the Rule Engine (it is now in jsnippets). It was created with a similar use case also in mind, so it would be easy to implement this change.