Options

How can I tackle the problem of naming examplesets?

pblack476pblack476 Member Posts: 83 Maven
edited February 2020 in Help
So I have a bunch of examplesets generated in a loop and I would like to organize them in the results tab by naming each one. They all just come out as generic ExampleSets inside IObjectCollection objects. Is there a way to organize the results tab in another way or at least name each exampleset of a collection?

Answers

  • Options
    sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
    @pblack476 ooh that's a good one. If it was not a loop it would be easy - just rename the OPERATOR as that's where the results tab name comes from:





    Hmm let me ponder a clever way to do this in a loop...

    Scott
  • Options
    pblack476pblack476 Member Posts: 83 Maven
    edited February 2020
    I usually do this as well. I also rename the first column of my dataset to have an easy to see "tag" on it. And That is the route I am taking so far. Renaming the first column with a macro tag that acts as an unique, useful identifier.

    But since there seems to be no real way of doing this I'll leave this here as a feature request.

    What are you supposed to do if your process generates hundreds or thousands of examplesets? Usually such a process is not very useful but it appears from time to time while building a larger and more efficient process.

  • Options
    sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
    @pblack476 I think the short answer at the moment is that if you're building that many ExampleSets in production, we recommend writing them into a database as tables. The results view was never designed to have hundreds of tabs (or hundreds of objects in a collection). You can name each table by the %{entry_name} macro and then go from there. You can write this in SQL directly, or use the In-Database Processing extension for a more drag-and-drop solution. A database of tables is exactly the same thing as a collection of ExampleSets. 

    The reason I say "at the moment" is that Engineering is completely rebuilding the repo structure and will release it soon. If you want to get a sneak peek, schedule a meeting with our UI/UX designer here: https://drift.me/egishijaku :wink:

    I'll file it as a feature request if you want, but again as I know the entire repo structure is being rebuilt from the ground-up, it's highly unlikely that it will be addressed. But I will ponder more...sometimes there are sneaky workarounds. 

    Scott
Sign In or Register to comment.