No difference between "polynominal" and "text" data type?

karl_trelakarl_trela Member Posts: 2 Contributor I
edited December 2018 in Help

Hi,

 

I checked the manual and this forum but I cannot find the answer.

 

Is there possibly no difference between the "text" and "polynominal" data types in RM?

 

I am asking, because I worked with quite an large data set and RM seems to save text variables the same way it does with polynominals: as categorical variables (or factors).

The metadata file gets really huge and this slows down RM a lot when loading and handling data.

 

Could this be true?

 

Best

Karl

Answers

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn

    The differences are not visible, but they are important.  On the back-end, there are many text processing operators that only work on data of type "text" and there are many normal operators that work on polynominal data types but not text.  So you do need to make sure that this is set correctly depending on what your downstream needs may be.

    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,503 RM Data Scientist

    @karl_trela,

    good catch. I think if i remember it correctly it is indeed at the moment the same implementation and the difference is on the usability. But this might change in future.

     

    Best,

    Martin

    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • t_liebet_liebe Member Posts: 14 Contributor I

    Hi guys,

     

    I am trying out some text processing models and they worked all for polynominal attributes.

    Now I want to try a new one and need to change the att into a text att. Converting the excel file into a csv didn't work. Can you help me with that ?

     

    Thanks in advance.

  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,635 Unicorn

    "Nominal to Text" operator will convert any attributes that you need.

     

    Brian T.
    Lindon Ventures 
    Data Science Consulting from Certified RapidMiner Experts
  • t_liebet_liebe Member Posts: 14 Contributor I

    Got it already but thank you anyway !

Sign In or Register to comment.