Skip to content

Speed up _create_topic_vectors by replacing DataFrame .loc with NumPy masking #2406

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.

Sign up for GitHub

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

Merged
merged 3 commits into from
Aug 17, 2025
Merged

Speed up _create_topic_vectors by replacing DataFrame .loc with NumPy masking #2406

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f050136
efficient indexing
jinsolp Aug 4, 2025
6a588d0
Merge branch 'MaartenGr:master' into efficient-create-topic-vec
jinsolp Aug 5, 2025
0b42b30
rm emptyline
jinsolp Aug 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions bertopic/_bertopic.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4315,11 +4315,11 @@ def _create_topic_vectors(
if embeddings is not None and documents is not None:
topic_embeddings = []
topics = documents.sort_values("Topic").Topic.unique()
topic_ids = documents["Topic"].values
doc_ids = documents["ID"].values.astype(int)
for topic in topics:
indices = documents.loc[documents.Topic == topic, "ID"].values
indices = [int(index) for index in indices]
topic_embedding = np.mean(embeddings[indices], axis=0)
topic_embeddings.append(topic_embedding)
mask = topic_ids == topic
topic_embeddings.append(embeddings[doc_ids[mask]].mean(axis=0))
self.topic_embeddings_ = np.array(topic_embeddings)

# Topic embeddings when merging topics
Expand Down
Loading