title	emoji	colorFrom	colorTo	sdk	app_port	pinned	license
Pneumonia Predictor	🚀	blue	indigo	docker	8501	false	gpl-3.0

Pneumonia Predictor

Documentation: https://datsudo.github.io/pneumonia-predictor/

Setup

Install the following prerequisites:

1. Python 3.12
2. Poetry (for dependency management)

Make sure that poetry is installed by running:

poetry --version

# Result must be:
# Poetry (version X.X.X)

Clone this repository:

git clone https://github.com/datsudo/pneumonia-predictor

Then install the project dependencies when you're inside the directory:

poetry install

Before working on this project, ALWAYS make sure that everytime you're working on this project, you're in the proper virtual environment created by poetry:

poetry shell

After that, setup the provided pre-commit hooks (to fix proper code styling for staged files before commiting) after cloning.

pre-commit install

Running

Sample usage

from pneumonia_predictor.backend.rf_active_smote import RfActiveSMOTE()

# Create an RFActiveSMOTE instance
rf_as = RfActiveSMOTE(X_train, y_train, X_test, y_test, "target_column_name")

# Training
rf_as.smote.train(n_iterations=5)

# Other options
rf_as.save("model_name")  # will save the model at ./saved_models/model_name.pkl

rf_as.display_results("acc")  # Print results [acc (Accuracy), min (Minority), maj (Majority), avg (Weighted Average)]

Web app

Open the app by running

streamlit run app.py

The web app will become accessible at http://localhost:8502.

License

This project is licensed under GNU General Public License v3.0.