Quickstart

Option A — Docker (recommended)

New to Docker? Start with Docker for beginners to install it and check your setup.

Build from source

git clone https://github.com/ranasl62/ehr-chronic-disease-risk-prediction.git
cd ehr-chronic-disease-risk-prediction
docker compose up --build
# stop: docker compose down

Pull published images (no local build)

Images are on Docker Hub (public). Pull either image alone, or use Compose:

docker pull ranasl62/ehr-risk-api:latest
docker pull ranasl62/ehr-risk-web:latest

git clone https://github.com/ranasl62/ehr-chronic-disease-risk-prediction.git
cd ehr-chronic-disease-risk-prediction
docker compose -f docker-compose.yml -f docker-compose.publish.yml pull
docker compose -f docker-compose.yml -f docker-compose.publish.yml up
# or: make researcher-up-pull

Defaults: ranasl62/ehr-risk-api:latest · ranasl62/ehr-risk-web:latest. Override with IMAGE_API / IMAGE_WEB in .env. Still clone the repo — Compose bind-mounts . for persistence. Full details: Docker Hub images.

Optional: cp .env.example .env (API key, ports). GPU / local overrides: docker-compose.override.example.yml.

data/, reports/, and model.pkl persist on the host. Demo data only — no PHI. For research and education only. Outputs are not clinical recommendations and are not intended for patient care. We are working toward broader general-purpose use in the future. · support@larucare.com.

Docker Hub images

Published by GitHub Actions on main / v* tags. Both repos are public.

docker pull ranasl62/ehr-risk-api:latest
docker pull ranasl62/ehr-risk-web:latest

Also tagged :sha-… and semver from release tags. See Docker Hub images for compose, pins, and troubleshooting.

First successful loop

1. Open Datasets → select the tiny longitudinal demo → Run health.
2. Open Train → start train (logreg is fine) → wait for succeeded.
3. Open Results / Analytics → inspect metrics and charts.
4. Optional: Generate SHAP · run leakage audit.
5. Open Predict → Fill medians → Predict.

Stop & clean

docker compose down                 # stop stack (keeps host data)
docker compose down --rmi local     # also remove project images
# optional host cleanup:
# rm -f model.pkl; rm -rf reports/runs data/uploads/*

Option B — Local Python + Angular

pip install -r requirements.txt && pip install -e .
PYTHONPATH=. uvicorn api.main:app --reload --port 8000

# another terminal
cd web && npm install && npm start
# UI http://127.0.0.1:4200 (proxies API)

Or CLI: ehr-ai start after editable install. Full install notes live in repository INSTALLATION.md.

Tests

PYTHONPATH=. pytest tests/ -q
cd web && npm run test:ci
# or: make test-all

Next reading