AP@K Calculate

[yanyijiang09]

Metrics/Python/ml_metrics/average_precision.py

Line 32 in 9a637ae

if p in actual and p not in predicted[:i]:

Hello: I notice that the reuslt from running apk([1, 1, 1], [1, 1, 1], 3) does not return 1. I wonder if it should be if p in actual and p in predicted[:i+1]? Thank you.

[maayanorner]

You are right - it's incorrect.
I believe this repo is not maintained for very long, but I would still deprecate it - I spent a few hours on that, and it did not help this error also propagated to a tutorial on the subject.

[fkurushin]

@yanyijiang09 @maayanorner I also noticed that in the return statement the authors divide precision by the number of actual documents, but they should divide precision by the number of relevant documents.

https://en.wikipedia.org/w/index.php?title=Information_retrieval&oldid=793358396#Average_precision