Closed ai-honzik closed 2 years ago
You are absolutely correct. != 0
is what I intended. But using the == 0
feature should result in the same classification performance on most ML algorithms. And the count of nonzero sections is derivable from the features. So I think I'll just bring the comment in line with the existing code and generated data. Thanks!
Hi, I was going through the code and assumed (given by the comment above) that line 168 in __init__.py should have been
sum(1 for s in sections if s['size'] != 0)
. https://github.com/elastic/ember/blob/4dee42918694d72d319e731940755146a71f5c6c/ember/features.py#L168Regards.