7 • Advanced topics II : sharing and deploying
Powerful things that are worth pointing out.
Options for sharing your data app
There are a few ways to share your app with people, and it will be worth doing a search for current options, as these will change with the times.
Streamlit Cloud
The easiest way to deploy is through Streamlit Cloud . The only requirement is that you have a GitHub account - the GitHub repo then acts as the source for the app being run by Streamlit Cloud. Full instructions on how to do this are here - they are very easy to follow!
For small apps, hosting of your app is free. If data usage or visit counts are high, you might need to start a paid tier.
Heroku
Heroku is another cloud service provider with special channels for deploying Streamlit apps. For current instructions on how to do this, it will be best to do a search as those instructions will update often.
Sharing Python code
If the person you are sharing with can run Python, you can just share your repository (or a folder of the script files). They can then install Streamlit, and run the app locally. This is a good option if you have data which needs to remain private.
Streamlit Desktop App
This is fairly new in development, but it may be possible to create an executable desktop app of Streamlit . Again, look up current articles and threads on the best way to do this.
Some final things to highlight!
As of 2023, Streamlit is still rapidly acquiring new functionality. Here are some things that are very powerful for creating a more sophisticated app:
Session state
The session state allows you to store variables across runs, and is generally useful for organising data across your app. Learn how to use it here. .
Interactive dataframes
Being able to edit or add data in the native app is a useful feature. Check here for current news on how this is done in Streamlit (keep in mind this link might change one day - if it doesn’t work, do a search instead!)
Data linkage
Streamlit is built to work with the APIs of major data infrastructure and software , including SQLite, AWS and MongoDB.