Building interactive data dashboards with Streamlit in Positron

Install Positron

First, download Positron from the official download page.

Install Python

If you don’t already have Python installed, you can follow the recommended instructions to set up a new version on your computer. In my case, I used pyenv to install version 3.13.1, although if you have already installed uv, we also support using it for managing virtual environments or even bootstrapping Python.

Open the Project Folder

Open up Positron, and from the Welcome Page, click “New from Git”.

Then paste the GitHub repository URL in the dialog box:

https://github.com/posit-marketing/streamlit-example.git

Save it in your desired directory, then open the cloned repository. You’ll now be in the project folder. On the left activity bar, you can see the project files in the File Explorer, which you can view and open:

streamlit-example/
├── app.py            # Main Streamlit application
├── original-app.py   # Original Streamlit application
├── trauma.geojson    # Hospital location data
├── requirements.txt  # Python dependencies
├── .gitignore        # Version control exclusions
└── README.md         # Project documentation

Quick app overview

The original-app.py code, available on GitHub, contains several key Streamlit components:

• A state filter using st.selectbox()
• The @st.cache_data decorator to optimize data loading
• Healthcare metrics displayed with st.metric()
• A histogram using st.bar_chart()
• Streamlit’s built-in map with st.map()

To take a closer look at the data, run this code in the Positron Console or a new Python script:

import geopandas as gp
trauma = gp.read_file("trauma.geojson")

This loads the trauma.geojson data into your Positron session, populating the Session pane in the Secondary sidebar. Click on the variables to preview their contents.

Use the table icon (or %view trauma in the console) to open the ‘trauma’ dataframe in the Data Explorer. Here, you can examine summary statistics, sort, filter, and browse the data in a spreadsheet-like view. Once you’re done, close the Data Explorer and return to original-app.py.

Previewing the app

You have everything you need to run your Streamlit app! Open original-app.py and click the “Play” button to see a preview in the Viewer pane.

You can interact with the app directly in the Viewer pane, or click the “pop out” button to open it in your default browser.

To enable live reloading, save app.py after making an edit. The app will prompt you to “Reload” or “Always Reload.” Click “Always Reload” to enable automatic reloading for this session.

For instantaneous updates anytime you open this folder, configure your Streamlit config.toml file. From your Streamlit app’s root directory (where app.py is), run in the Terminal:

mkdir -p .streamlit
nano .streamlit/config.toml

Then, add these settings to the nano editor:

[server]
runOnSave = true

Press Ctrl + O (to save), Enter (to confirm filename), then Ctrl + X (to exit). Alternatively, manually create the .streamlit folder and config.toml file, then paste and save the configuration. Now, your preview will automatically update whenever you save changes in app.py. Try it out by editing the title of the dashboard.

Debugging the app

Positron includes robust debugging features for app development. Set breakpoints in the editor margin, click the “Play” button, and then select “Debug Streamlit App in Terminal.” Positron will run the app in a dedicated Terminal tab, open the app URL in the Viewer pane, and launch it in debug mode.

Code assistance with Positron Assistant (Preview)

When I first created this dashboard (two years ago!), I focused on demonstrating deployment to Posit Connect. For this tutorial, I wanted to improve its features and functionality.

Positron Assistant (a preview feature in Positron 2025.07.0-204+) provides LLM integration for chat (with Anthropic) and inline completions (with GitHub Copilot). You can configure it by following thePositron Assistant documentation.

Once configured, open it from the Activity Bar. It automatically detects app.py and relevant environment variables.

I wanted a quick overview of key metrics for each state, such as the number of helipads. I asked Positron Assistant, “Can you add a row of metrics to the dashboard using st.metric? I would like the three metrics to be number of hospitals, number of level I trauma centers, and number of helipads, with the data filtered by whatever is chosen in st.filter”. Positron Assistant performed the necessary calculations and provided complete code snippets.

You have a few options for adding the suggestions: edit and copy the code from the chat, insert the code into a new file, or, what I found most helpful, insert it right at your cursor, as Positron Assistant guides you to the correct spot.

I then built out the rest of the dashboard, adding more metrics and leveraging Positron Assistant for further suggestions that shaped the final design.