Sepsis Prediction Methodology

This repository contains a rewritten version of the code for our submission (Team Name: Can I get your Signature?) to the PhysioNet 2019 challenge.

Getting Started

Once the repo has been cloned locally, setup a python environment with python==3.7 and run pip install -r requirements.txt.

You then need to add the project root directory to your virtualenv python's path. This can be done by adding the location of the root folder into the site_packages.pth directory. (See Arjen P. De Vries answer here: https://stackoverflow.com/questions/4757178/how-do-you-set-your-pythonpath-in-an-already-created-virtualenv/47184788#47184788).

Create a folder /data/raw, the data/ folder should be made a symlink if you wish to store the large data files elsewhere.

Run the following:

1. python src/data/get_data/download.py To download the raw .psv files to /data/raw
2. python src/data/get_data/convert_data.py To convert the downloaded data into a pandas dataframe (for easy analysis) and a TimeSeriesDataset (for fast operations).

You are then ready to go! Check /notebooks/examples/prediction.ipynb for an intro to the basic prediction methods and the functions used to generate the features. Then either follow the example and use a notebook to build your own models, or do something similar to that seen in src/model/examples/train_{MODEL_TYPE}.py.

More functionality will be added soon!