Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Uniform testing #558

Open
Liikt opened this issue Dec 24, 2018 · 3 comments
Open

Uniform testing #558

Liikt opened this issue Dec 24, 2018 · 3 comments
Labels
Discussion This is open for a discussion. General Problem This is a problem in the archive or an implementation.

Comments

@Liikt
Copy link
Contributor

Liikt commented Dec 24, 2018

While looking through the code of PR #542 we came across the problem of sometimes not having proper test cases. In this PR random numbers are generated and you can't really make out if the computation actually succeeded or not.

Some implementations do have that but some don't and I think this is something we should consider.

I'll leave the issue here for discussions and ideas.

@leios leios added Problem This is a problem in the archive or an implementation. General Discussion This is open for a discussion. labels Dec 24, 2018
@leios
Copy link
Member

leios commented Dec 24, 2018

This is closely related to #320. I agree, we need to create more rigorous tests for most methods. This is part of the upcoming revisions that I am working on now. That said, I would love to have some discussion on what tests are most worthwhile for the methods we currently have.

@leios
Copy link
Member

leios commented Apr 10, 2019

After a discussion on stream, we realized that it wasn't clear what was meant by "uniform testing" I had naively assumed it meant that most main() functions would include a test (if possible), like so:

function main()
    init()
    run_method()

    if(test_passed())
        println("This method works!")
    else
        println("This method doesn't work!")
    end
end

Please let me know how you would like for this testing to happen.

@Liikt
Copy link
Contributor Author

Liikt commented Apr 11, 2019

Yes basically that we have (where it works) a defined input with defined output. If randomness is involved than we may need a threshold but I think the most algorithms don't rely on randomness so something like you described would be possible.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Discussion This is open for a discussion. General Problem This is a problem in the archive or an implementation.
Projects
None yet
Development

No branches or pull requests

2 participants