Question: Select Attributes and classify them

kloriklori Member Posts: 3 Contributor I
edited December 2019 in Help

Hi everyone : )

I am new using rapidminer and I hope somebody can help me please?

I am currently working on a logistic regression workflow which retrieves the data directly from the DWH.

Before I used read - csv where I could classify the information of the file and deselect attributes.


But if I want to get the data fromn the DWH i dont know how to classify the data and select the attributes : (

To specifiy what I want:


I want to choose which variables should by analyized (deselection variables) and I want to specifiy the variables for "polynomial" and one as "binomial" and classify it also as label? Can anyone help me please?




Here an example of my WFexample.jpg



  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,508 RM Data Scientist

    Hi Klori,


    you can either use a SQL query in the Read Database operator to select the cols you would like to use. Otherwise you can do this in RapidMiner using the Select attributes operator.

    You pretty much copy/paste you upper stream to the lower DWH operator.


    One thing: You apply the model comming out of X-Val on the data it is learned on. Please be aware that this is no good idea if you would like to analyse the data later on, because of over training. If you really need a scored data set, please use the X-Prediction operator.



    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    yyhuangyyhuang Administrator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 364 RM Data Scientist

    You can use 'set role' operator to set the target role as 'label' for the target varible of your prediction. 

    Some data type convertion operotors could be used for your 'specifiy the variables for "polynomial" and one as "binomial"  '


    Use 'subset' of attributes as your selected/not selected filter in 'Select Attrbutes', and you can also choose 'invert selection'  




Sign In or Register to comment.