🦉🦉   WOOT WOOT!   RAPIDMINER WISDOM 2020 EARLY BIRD REGISTRATION ENDS FRIDAY DEC 13!   REGISTER NOW!   🦉🦉

How to include an id# in auto model predict result without it contributing to input.

donsdons Member Posts: 4 Newbie
edited November 9 in Help
I've tried making it an id or label.  It doesn't appear in the result column unless I make it an input, in which case it contributes to the input which I don' want.
Tagged:

Best Answer

Answers

  • lionelderkrikorlionelderkrikor Moderator, RapidMiner Certified Analyst, Member Posts: 911   Unicorn
    Hi @dons,

    I'm not sure to understand.
    But if we take example on the Iris dataset : 
     1/ Check the Id column in order the id  is taken into account



     2/ Execute Automodel
    3/ Open a process (when the execution is completed).
    4/ Go to Load and Process Data



    5/ GO to Preprocessing --> Define Target ?

    6/ In the Define Target (Set Role) operator set your id column as id in the parameters.

    This way, your id column/attribute will appear as "id" in the results tables.

    Hope this helps,

    Regards,

    Lionel



  • donsdons Member Posts: 4 Newbie
    Thank you Lionel.  After Predict runs, I click on General, Data, and Open Process. Then as you indicated, Preprocessing, and Define Target. Then Set Role  example set output.
    I highlight the out circle on the far right and click on the id#.  The Role box is blank, and I don't see a way to set it to id. Don
  • donsdons Member Posts: 4 Newbie
    Thank you Lionel.  Please check this,  After Predict runs, I clicked on General, Data, and Open Process. Then as you showed, Load and Process Data, Preprocessing, Define Target. In the upper right Actions is picked. On the far right, I click on Set Role examples set output
  • donsdons Member Posts: 4 Newbie
    Thank you Lionel for your detailed instructions.
    Don
    lionelderkrikor
  • lionelderkrikorlionelderkrikor Moderator, RapidMiner Certified Analyst, Member Posts: 911   Unicorn
    You're welcome, Don.

    Regards,

    Lionel
Sign In or Register to comment.