Nominal to Binominal
operator has a "transform binominal" checkbox which does exactly that.
You don't have to transfer the binominal attribute to nominal first. It works even for binominal attributes on input.
Question
I have imported a dataset in which the gender (m/f) has been interpreted as "binominal". If I am not mistaken, this means that there is a mapping true/false is assigned to the two values m/f.
Since I need to find association rules, this is not correct.
My question is: is it possible to convert a binominal field to nominal and then expand it again to two nominal attributes (one for male and the other for female) in order to be correctly used with the "FP-Growth" and "Create Association Rules" operators?
Solution
Nominal to Binominal
operator has a "transform binominal" checkbox which does exactly that.
You don't have to transfer the binominal attribute to nominal first. It works even for binominal attributes on input.