This is a minimal implementation of NFM framework in the paper Neural Frailty Machine: Beyond proportional hazard assumption in neural survival regressions

The python module nfm contains all functionalities of nfm. For demo purpose, we provide two scripts over the metabric dataset (included in the data repository) run

python experiment_metabric_pf.py

or

python experiment_metabric_fn.py

For evaluations on other datasets, check the usage of nfm.datasets.SurvivalDataset. Specifically, replace SurvivalDataset.metabric with other dataset configurations, and the rest of the codes require almost no change.

For reproduction of synthetic experiments, check experiment_synthesis.py Note that the preprocessing routine in the paper is slightly different from the standard routine in pycox package