Skip to content

Add more "synthetic" datasets in the contrib.learn #5314

Closed

Description

@z-a-f

The feature request is as in the title -- having synthetic datasets for demo purposes, and to see how different algorithms behave would be beneficial. It is fairly straightforward to generate two-dimensional multi-class set for a lot of synthetic problems.

Activity

  1. aselle commented on Nov 1, 2016

    @aselle
    Contributor

    @martinwicke @ilblackdragon, do you have any comments on this. I could mark this contributions welcome.

  2. vrv commented on Nov 1, 2016

    @vrv

    See #5315

  3. changed the title [-][Enhancement] It would be good to have "synthetic" datasets in the `contrib.learn`[/-] [+]Add more "synthetic" datasets in the `contrib.learn`[/+] on Nov 2, 2016
  4. z-a-f commented on Nov 7, 2016

    @z-a-f
    Author

    This should be reopened -- there was accidental contamination which caused the commit comment include the reference to the current issue. The actual solution to the current issue is #5315

  5. reopened this on Nov 7, 2016
  6. reopened this on Nov 10, 2016
  7. z-a-f commented on Nov 10, 2016

    @z-a-f
    Author

    Sorry, accidentally closed.

  8. added a commit that references this issue on Nov 14, 2016
    cb2ef58
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions