'HRLKNNClassifier:trainClassifierWithMatrix:' & 'HRLKNNClassifier:calculateClassificationAccuracyUsingMatrix:' should return an error if the classifier is not trained

HeartRateLearning / HRLAlgorithms

(DEPRECATED) Wrapper around LearnKit

MIT License

1 stars 1 forks source link

Closed indisoluble closed 7 years ago

indisoluble commented 7 years ago

There are a few options:

Return error + nil
Only nil
Return 0 (or any other value) and specify in the documentation that, until the classifier is trained for the first time, the returned values are wrong.

indisoluble commented 7 years ago

Thanks to changes done in commit: 013cfad78d7d3bff7872568934138a6f12b03888, this issue no longer applies.