How to branch based on attribute type in Rapidminer?

Question

I would like to do a bunch of statistics on a huge data set, so I'm looping through all the attributes with Loop Attributes. In the operator, I have an Aggregation operator, which doesn't work in itself of course. The solution would be doing a Branch on different data types.

Does anybody know, how to do that?

Answer 1

I don't understand what your process is trying to achieve. Do you mean you have many attributes that you want to aggregate based on their type. So you might have 300 attributes in total of which 100 are numeric, 100 are binominal and 100 are polynominal. Do you then want to aggregate the 100 numerical attributes into an average for each example, count the 100 binominals and perhaps find the mode of the polynominals? The end result would be three attributes summarizing each example.