CardioVar_v1

This bundle runs out of the box using a built-in BaselineModel if no trained model is provided. You can still point to your own *_best_model.joblib later via env vars.

1) Create environment

Windows PowerShell

cd CardioVarApp_Working_20250907_070933
python -m venv .venv
.\.venv\Scripts\Activate.ps1
pip install -r backend\requirements.txt

macOS/Linux

cd CardioVarApp_Working_20250907_070933
python3 -m venv .venv
source .venv/bin/activate
pip install -r backend/requirements.txt

2) (Optional) Use your trained model

$env:MODEL_PATH="C:\path\to\rf_best_model.joblib"   # optional
$env:FEATURES_PATH="C:\path\to\features_pruned.txt" # optional
$env:THRESHOLD="0.5"
$env:LOW_THRESH="0.3"
$env:HIGH_THRESH="0.7"
$env:CLASS_MAP="0:not candidate,1:candidate"

If MODEL_PATH cannot be loaded, the server will automatically use a BaselineModel so you can test immediately.

3) Run the server

python -m uvicorn backend.app:app --host 127.0.0.1 --port 8000
# Then open http://localhost:8000   (API docs at http://localhost:8000/docs)

4) Test with the provided sample

• In the web UI, click Upload & Run CardioVar → choose sample\example_variants.csv → Run CardioVar.
• You will see a preview table and a Download results (CSV) button.