Metadata-Version: 2.4
Name: actor_finder
Version: 0.1.0
Summary: Doc2Vec 학습 + 클러스터 진단(실루엣, 덴드로그램) 유틸
Author: actor_finder contributors
License: MIT
Project-URL: Homepage, https://pypi.org/project/actor_finder/
Project-URL: Repository, https://github.com/smhrdGit/New_value_Analysis.git
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy>=1.21
Requires-Dist: pandas>=1.5
Requires-Dist: gensim>=4.3
Requires-Dist: scikit-learn>=1.2
Requires-Dist: matplotlib>=3.7
Requires-Dist: scipy>=1.10
Dynamic: license-file

# actor_finder

Doc2Vec 학습 + 군집(클러스터) 진단을 위해 자주 쓰는 함수들을 묶은 유틸 라이브러리입니다.

## 로컬 설치(개발 모드)
```bash
pip install -U pip
pip install -e .
```

## 빠른 사용 예시
```python
from actor_finder import train_doc2vec, agglomerative_silhouette

# df["tagged_review"] : 토큰 리스트 컬럼
model, vectors, tagged = train_doc2vec(df, token_col="tagged_review")
df["vector"] = vectors

scores, best_n = agglomerative_silhouette(df, vector_col="vector", n_cluster_range=range(2, 10))
print(best_n, max(scores))
```
