Best Answer

• Options

  varunm1 Moderator, Member Posts: 1,207 Unicorn

  February 2020 edited February 2020 Solution Accepted

  Hello @raw160107
  
  
  You don't need to connect the output ports of write CSV. You can just define the name and location where you want to store and run the process. This works fine for split validation.
  
  In the case of cross-validation, this doesn't work as there are multiple trains and test sets based on a number of folds. One way that I do it is by using macros.
  
  When you give the name for CSV file, you need to specify a macro %{execution_count} . This will help store training and test separately and also let you be clear on which fold they belong to.
  
  I attached sample process .rmp file in this thread, please go through the csv file parameter of write CSV to see how I used macros. To see the attached process, you need to download it from this thread and then in rapidminer studio go to FILE --> IMPORT Process and navigate it to this file.
  
  The train and test file names in write CSV of cross-validation are named as mentioned below. You can use any name but in the end you should add _%{execution_count}
  
  Cross_Train_Fold_%{execution_count}
  Cross_Test_Fold_%{execution_count}
  
  Once the process is run it will create files names as Cross_Train_Fold_1.csv, Cross_Train_Fold_2, .... based on the number of folds in CV.
  
  Please let us know if you need more information.
  
  

  Process_Storing_Train_Test.rmp 15.1K

  Regards,
  Varun
  https://www.varunmandalapu.com/
  
  

  Be Safe. Follow precautions and Maintain Social Distancing

  4