Problem with Performance (Attributes count)

vikturek

Hello,

I want to use the operator Performance (Attributes count) for decision which neural network model should I use. This operator requires as input an Example set and "performance vector" which should be optional according to all sources (for example official tutorial). However my studio is telling me: "Mandatory input missing at port Performance.performance" which is obviously incorrect. I don't know what should be conected to this input and why it is mandatory. My only goal is to calculate the count of attributes in example set, that's all.

 

Note: I tried an older version of RapidMiner (5.3) and the newest version 7.1 on Linux and Windows. The non-working situation can be seen on the attached picture and you can compare it with official tutorial version at http://docs.rapidminer.com/studio/operators/validation/performance/predictive/performance_attribute_count.html

MartinLiebig

Dear Viturek,

 

i am not sure about the problem with the operator, but you can use Extract Performance and select number of attributes to get the same effect.

 

~Martin