apd-crs is a Python package for survival analysis with cure rate. I.e. time-to-event analysis using a dataset for which outcomes of a fraction of the dataset are unknown (censored), and for which some individuals never experience the event. For example:
a medical trial where time since diagnosis is measured and some participants drop out during the trial period, some of which are cured.
a manufacturing dataset where time-to-failure since maintenance is measured, and some equipment takes so long to fail that it “never sees” the event.
Currently there are three methods available for estimating the cure probability per covariate:
using the selected completely at random (SCAR) assumption,
using a Hard EM algorithm.
The overall survivor function is fitted by assuming that the lifetime of a susceptible (non-cured) individual follows a proportional hazards model, and the baseline hazard function is given via Weibull distribution.
Since the cure probability is fitted independently of the times to event/censoring, this package can also be used for PU learning. I.e. binary classification with positive and unlabeled data.
apd-crs is a work in progress and maintained by a team of researchers and developers from Aimpoint Digital.
Using pip by running:
pip install apd-crs
Found typos? Interested in new functionalities? Please submit bug reports and feature requests on our GitHub repository.