Nitin Madnani comments

Results 127 comments of


                                            Nitin Madnani

Use black to standardize code formatting

Makes sense. Thank you so much!

Kappa computation when predicted scores are on a different scale

> if the range is greater than certain threshold. what does `range` refer to exactly? The range of the predicted scores?

Kappa computation when predicted scores are on a different scale

Ah, I see. And what sort of threshold would you use?

Kappa computation when predicted scores are on a different scale

I don't think we want to go for the worst case number like that since that can still lead to memory errors. I feel like this threshold should depend on...

Kappa computation when predicted scores are on a different scale

Is there a realistic scenario where the ranges diverge by more than that and kappa will still be meaningful?

Don't update user's jupyter history when notebooks are executed

@mulhod ☝️ ?

Look into parallelization

Output file or directory in rsmpredict

Actually, now that I think a bit more about it, I think we should strive for consistency. So, here's an alternative I'd prefer: - Make `rsmpredict` also use an output...

Output file or directory in rsmpredict

Hmm, yes I can see how that can be quite annoying. I like your suggestion! 👍

Rewrite docstrings to match scikit-learn Style

See also [this issue](https://github.com/EducationalTestingService/skll/issues/647) for SKLL.