FRAILTY-BASED COMPETING RISKS MODEL FOR MULTIVARIATE SURVIVAL DATA

A Python package for frailty-based multivariate survival data analysis with competing risks.

Danielle Poleg, Malka Gorfine 2023

Installation

pip install PyFCR

A pre-requisite is having python 3.10 installed on windows and python 3.11 on MAC

Quick Start

This package has 2 run types:

Simulation

Simulation is used to examine the model's performance. To start with a simple simulation, run the following code:

from pyfcr import FCR, Config, RunMode, FCRType

if __name__ == "__main__":
    config = Config(run_type=RunMode.SIMULATION, fcr_type=FCRType.BIOMETRICS)
    fcr = FCR(config)
    fcr.run()
    fcr.print_summary()

To perform your own simulation with non-default parameters, add your configurations to the Config object.

Analysis

Analysis is used to evaluate the model on an existing dataset. The dataset should consist of 3 different csv files:

• deltas.csv - the event types data. Should be a file of size: n_clusters * n_members * n_competing_risks
• X.csv - the event times data. Should be a file of size: n_clusters * n_members * n_covariates
• Z.csv - the covariates data. Should be a file of size: n_clusters * n_members

from pyfcr import FCR, Config, RunMode, FCRType

if __name__ == "__main__":
    config = Config(run_type=RunMode.ANALYSIS, fcr_type=FCRType.BIOMETRICS, data_path='<path_to_data_files>')
    fcr = FCR(config)
    fcr.run()
    fcr.print_summary()

How to Contribute

1. Open Github issues to suggest new features or to report bugs\errors
2. Contact Danielle if you want to add a usage example to the documentation
3. If you want to become a developer (thank you, we appreciate it!) - please contact Danielle for developers' on-boarding

Danielle Poleg: daniellepoleg@gmail.com