Metadata-Version: 2.1
Name: survshap
Version: 0.4.0
Summary: Implementation of the SurvSHAP(t) explanation method for time-dependent explainability of machine learning survival models
Project-URL: Homepage, https://github.com/MI2DataLab/survshap
Project-URL: Bug Tracker, https://github.com/MI2DataLab/survshap/issues
Author: Mikołaj Spytek, Hubert Baniecki, Przemysław Biecek
Author-email: Mateusz Krzyziński <krzyzinskimateusz23@gmail.com>
License-File: LICENSE
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.7
Requires-Dist: numpy>=1.20.3
Requires-Dist: pandas>=1.2.5
Requires-Dist: plotly>=5.1.0
Requires-Dist: scikit-survival>=0.17.2
Requires-Dist: scipy>=1.6.3
Requires-Dist: shap>0.41.0
Requires-Dist: statsmodels>=0.13.2
Requires-Dist: tqdm>=4.61.2
Description-Content-Type: text/markdown

# survshap

<!-- badges: start -->
<!-- badges: end -->

## Overview 
The `survshap` package contains an implementation of the **SurvSHAP(t)** method, the first time-dependent explanation method for interpreting survival black-box models. It is based on SHapley Additive exPlanations (SHAP) but extends it to the time-dependent setting of survival analysis. SurvSHAP(t) is able to detect time-dependent variable effects and its aggregation determines the local variable importance.

Read more about SurvSHAP(t) in [our paper](https://doi.org/10.1016/j.knosys.2022.110234).

## Installation
```python
pip install survshap
```

## Citation
If you use this package, please cite our paper:
    
```bib
@article{survshap,
    title = {SurvSHAP(t): Time-dependent explanations of machine learning survival models},
    journal = {Knowledge-Based Systems},
    volume = {262},
    pages = {110234},
    year = {2023},
    issn = {0950-7051}
    }
```
