RapidMiner 9.8 Beta is now available

Be one of the first to get your hands on the new features. More details and downloads here:

GET RAPIDMINER 9.8 BETA

how to make summary tables from output data

zvisnozvisno Member Posts: 5 Contributor II
edited November 2018 in Help
hey

need help

i need to write a summary table from the result i receive from next process:

the process are:
1) i read data from AML
2) then I optimize it by generation (YAGGA)
3) validate with help of naive bayes
4) apply model and perform it

I have some inputs and would like to receive it in a file (an example performance vector data)

could someone tell me how  to do it with meanings (operators) of RM on example

thanks

Answers

  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Hi,
    I'm sorry, but I cannot comprehend what you are going to do. Could you repeat it and be more detailed?

    Greetings,
      Sebastian
  • zvisnozvisno Member Posts: 5 Contributor II
    I have to implement some process on data set.

    I have a dataset. I perform naive bayes algorithm for that dataset. There is a result (result overview mode) I have some data about that process and I wont to save that data and tables to file.

    Is it possible to do it with repidminer automatically  and how?

  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Yes,
    of course this is possible. Take the IOObjectWriter for writing every possible result object into a file. Just select the type of object you want to save and specify the file.

    Greetings,
      Sebastian
  • zvisnozvisno Member Posts: 5 Contributor II
    Hello

    It's clear how to write data into a file with a help of write-operators

    but could someone tell me how to write resultfile in 5.0 . I feel like it's not just to put a path, because it doesn't work in that way.
    In 4,6 version you should just to take an operator ResultWriter and indicate a path but what about 5.0?

    Thank you
  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Hi,
    there are two ways in RapidMiner 5.0. You could either store the data in a 4.x wise manner directly in files, or you could simply put them into the repository using the store operator. The latter is preferable, because it has many advantages. It's faster and let's you manage your project data directly from within RapidMiner. But the greatest advantage is, that the meta data of your data is stored, too. So only if you load a data set from repository, all the meta data transformation will be possible, which gives you so many help during process design. I really would recommend using the repository. I bet you will never want to work without the meta data again...
    The old way, which is still necessary for exporting data to another application is still possible. You find the operators in the import and export branch.

    Greetings,
      Sebastian
Sign In or Register to comment.