- Displays full dataframes with cleaned object types for safe rendering.
- Builds graphs using TF-IDF vectorization on sanitized text inputs.
- Handles edge cases like empty tables or insufficient text data.
Install via pip:
pip install streamlit pandas networkx plotly scikit-learn sqlalchemy pymongo pymysql psycopg2 pymssql oracledb
- Streamlit launches with streamlit run dataviewer.py. No additional configuration needed.
- Graphs require at least two non-empty text entries, large datasets may slow vectorization.
- MongoDB limits to 5000 documents.
- No query customization beyond full table loads. Assumes text columns are object/string dtype.