🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

label role for more than one attribute

sgenzersgenzer 12Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959  Community Manager
edited November 2018 in Help
How do I get more than one attribute to have a label role without going into the data editor?  If I use the Set Role operator, all I do is set one attribute, and if I try to do another, the prior one gets replaced.

Thanks.

Scott

Answers

  • David_ADavid_A Administrator, Moderator, Employee, RMResearcher, Member Posts: 277  RM Research
    Hi Scott,

    The label Role, as well as the ID for example, is a unique role.This is normally very useful otherwise making joins or validation would get very complicated.
    What you can to is to set the role of your first label to 'regular' if you want to keep it when adding another label

    Best,
    David
  • JEdwardJEdward RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 578   Unicorn
    You can actually free-type when entering in the values so you can have attributes with the role label, label2, label3, etc.

    Use the operator Loop Labels to cycle through those labels for different predictions.  (See model documentation for a sample process)
  • sgenzersgenzer 12Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959  Community Manager
    thank you both.  I know you can have more than one label attribute because you can see it as a result when you run certain processes that I did not create (e.g. if you use the Aylien Sentiment Analysis operator).  Yes I know I can do it manually (i.e. via the data editor) but was looking for a simpler, more automatic way to do it.  I will try the looping and see how that goes.

    Scott
Sign In or Register to comment.