GTPCNZ Interactive Reporting & Visualization Suite

Canonical GitHub Pages front door: https://edithatogo.github.io/gtpcnz/

Hugging Face interactive lab: https://edithatogo-gtpcnz-dashboard.hf.space/

Legacy Streamlit compatibility URL: https://gtpcnz.streamlit.app/

This is a public-data anchored benchmark and educational explainer. It is not linked-data calibrated and not a patient-level forecast. The Dash lab is the canonical interactive surface; the Streamlit app is retained only for legacy compatibility while the migration is checked. It should not be used to claim precise fiscal savings, ED reductions, hospital-demand reductions, workforce effects, implementation impacts, or causal effects.

This project includes a reproducible reporting layer and an interactive dashboard designed to make the primary care funding model accessible to both policy experts and non-specialist readers.

The GitHub Pages site now also has a canonical site map and release manifest at docs/public-site/site-map-and-release-manifest-v1.8.4.md so the public bundle has one explicit index.

1. Quarto Report: reports/primary_care_architecture.qmd

The Quarto report is a reproducible document that combines technical thesis writing with live Python data analysis.

How to Render

To generate the HTML or PDF version, ensure you have Quarto installed and run:

quarto render reports/primary_care_architecture.qmd

Design Philosophy

  • Narrative first: Explains the “Game Theory” of healthcare using relatable metaphors (e.g., video games, subscription services).
  • Traceable model outputs: Pulls directly from outputs/full-parameterised-summary-results-v1.7.0.csv.

2. Dash model lab: dash_app/app.py

The future interactive dashboard target is a Plotly Dash app deployed to Hugging Face Spaces. GitHub Pages remains the polished public front door.

How to Run

Use the repo-local Prefix.dev Pixi wrapper:

python scripts/bootstrap_prefix_pixi.py
python scripts/run_pixi.py run dash

If the bare pixi command resolves to another executable, keep using the wrapper:

python scripts/run_pixi.py --version

Features

  • Scenario comparison with Plotly charts, table fallback, interpretation, and CSV download.
  • Bounded uncertainty, stock-flow, agent-lens, and educational simulation views.
  • Public caveat strip and GitHub/Hugging Face topology links, with Streamlit labelled as legacy compatibility where shown.
  • Hugging Face Space packaging under dash_app/.

3. Streamlit compatibility dashboard: streamlit_app.py

GTPCNZ retains this Streamlit app as a legacy compatibility surface for comparing migration parity against the Dash lab.

How to Run

Ensure the repo-local Pixi runtime is installed:

python scripts/bootstrap_prefix_pixi.py

Then launch the dashboard:

python scripts/run_pixi.py run -e dev streamlit run streamlit_app.py

Features

  • Interactive Sliders: Adjust Capitation vs. FFS weights.
  • Educational Tooltips: Instant definitions for complex healthcare terms.
  • Real-time Plotting: Shows qualitative changes in model-generated supply and hospital-pressure index values.
  • Compatibility Entry Point: streamlit_app.py remains the Streamlit Community Cloud entrypoint only while the Dash migration is checked.

Technical Standards

  • Modularity: The dashboard leverages the existing project data structure.
  • Accessibility: Content is simplified without losing technical rigor.
  • Transparency: All visualisations are generated from traceable project outputs.
  • Automated Testing: models/tests/test_app.py uses Streamlit’s native AppTest API; models/tests/test_dash_app.py covers the Dash shell.