pat2vec

Transforms single patient data into a single-row vector and batch patient data into a data matrix D ready for machine learning (binary classification).

Table of Contents

• Requirements
• Features
• Installation
• Usage
• Contributing
• License

Requirements:

• CogStack (cogstack_v8_lite)
• Elasticsearch
• MedCat
• CSV
• Multiprocessing
• OS
• Re
• Sys
• Warnings
• Numpy
• Pandas
• Pickle
• Scipy
• COGStats
• JSON
• Random
• Subprocess
• Time
• Traceback
• Datetime
• functools
• IO
• Pathlib
• Tqdm
• Colorama

Features:

• Single patient
• Batch patient
• Cohort search and creation
• Automated random controls
• Modular feature space selection

Installation

To install pat2vec, you can clone this repo:

Usage:

• Set paths

• Configure options

• Run all

Contributing

Contributions are welcome! Please see the contributing guidelines for more information.

License

This project is licensed under the MIT License - see the LICENSE file for details