Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update popularity bias at k #7

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Update popularity bias at k #7

wants to merge 2 commits into from

Conversation

nsbits
Copy link
Contributor

@nsbits nsbits commented Apr 5, 2022

Description

Line 118 in 'reclist/reclist/metrics/standard_metrics.py' (url)
The denominator should be len(p[:k)

def popularity_bias_at_k(y_preds, x_train, k=3):
    # estimate popularity from training data
    pop_map = collections.defaultdict(lambda : 0)
    num_interactions = 0
    for session in x_train:
        for event in session:
            pop_map[event] += 1
            num_interactions += 1
    # normalize popularity
    pop_map = {k:v/num_interactions for k,v in pop_map.items()}
    all_popularity = []
    for p in y_preds:
        average_pop = sum(pop_map.get(_, 0.0) for _ in p[:k]) / len(p) if len(p) > 0 else 0
        all_popularity.append(average_pop)
    return sum(all_popularity) / len(y_preds)

should not the len(_p) be len(_p[:k]) ?
we will be looking till the kth slice
This is what I think should be popularity_bias@k

def popularity_bias_at_k(y_preds, x_train, k=3):
    # estimate popularity from training data
    pop_map = collections.defaultdict(lambda : 0)
    num_interactions = 0
    for session in x_train:
        for event in session:
            pop_map[event] += 1
            num_interactions += 1
    # normalize popularity
    pop_map = {k:v/num_interactions for k,v in pop_map.items()}
    all_popularity = []
    for p in y_preds:
        average_pop = sum(pop_map.get(_, 0.0) for _ in p[:k]) / len(p[:k]) if len(p) > 0 else 0
        all_popularity.append(average_pop)
    return sum(all_popularity) / len(y_preds)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant