All the code for the Streamlit workshop for Omdena. Please be aware that some things are intentionally simplified because this is the code used for the Workshop.
For this workshop I used the famous Titanic Dataset to reduce the amount of preprocessing and focus on Streamlit features. The file is the titanic.csv file within the folder.
- Streamlit website: https://streamlit.io
- Request invite to Streamlit share: https://streamlit.io/sharing
- Streamlit share main page: https://share.streamlit.io
- Streamlit documentation: https://docs.streamlit.io/en/stable/installation.html
- Streamlit widgets: https://docs.streamlit.io/en/stable/main_concepts.html#widgets
- pandas==1.0.5
- seaborn==0.10.1
- streamlit==0.83.0
- plotly==4.11.0
- matplotlib==3.2.2
- numpy==1.17.0
- joblib==0.16.0
- scikit-learn==0.23.1
I used Python 3.7. Streamlit currently requires Python 3.6-3.8.
The .joblib file was generated using joblib and the code in this notebook.
Obs: locally the code was running fine, but in order to work on Streamlit Share I also had to change the return of this function: https://github.com/victoraccete/omdena_streamlit_workshop/blob/c9330fd2c4a8b62a07d21a35386a421497c159dc/app.py#L53-L56
The requirements.txt file is necessary for streamlit share.
This is the link to access this app on Streamlit Share.