| import numpy as np | |
| import topicwizard | |
| from sklearn.datasets import fetch_20newsgroups | |
| from sklearn.feature_extraction.text import CountVectorizer | |
| from turftopic import KeyNMF | |
| print("Fetching data") | |
| newsgroups = fetch_20newsgroups( | |
| subset="all", | |
| remove=("headers", "footers", "quotes"), | |
| ) | |
| texts = newsgroups.data | |
| labels = list(np.array(newsgroups.target_names)[newsgroups.target]) | |
| model = KeyNMF( | |
| 20, | |
| vectorizer=CountVectorizer( | |
| stop_words="english", | |
| max_features=8000, | |
| ngram_range=(1, 2), | |
| ), | |
| ) | |
| topic_data = model.prepare_topic_data(texts) | |
| topicwizard.easy_deploy(topic_data, dest_dir=".") | |