yzhao062/pyod

knn metric cosine ValueError: Unrecognized metric 'cosine'

Open

#221 opened on Aug 19, 2020

View on GitHub
 (3 comments) (0 reactions) (1 assignee)Python (7,762 stars) (1,308 forks)batch import
bughelp wanted

Description

clf = KNN(metric="cosine")
clf.fit(train_data)

results in

sklearn/neighbors/_binary_tree.pxi in sklearn.neighbors._ball_tree.BinaryTree.__init__()

sklearn/neighbors/_dist_metrics.pyx in sklearn.neighbors._dist_metrics.DistanceMetric.get_metric()

ValueError: Unrecognized metric 'cosine'

Also brute and kd_tree don't support cosine, meaning that cosine is not supported by any algorithm.

Contributor guide

Tech stack
pythonscikit learn
Domain
machine learning
Issue type
bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
blocked
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
Pythonscikit learn
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40
Research direction
The issue is that the 'cosine' metric is not recognized by sklearn's BallTree or KDTree. The error trace indicates it stems from DistanceMetric.get metric. To fix, one could add 'cosine' to the list of supported metrics in pyod's KNN implementation, possibly by mapping it to a valid metric or computing cosine distance manually. Relevant files include pyod/models/knn.py or pyod/utils/utility.py.