sparsesurv [1] is a toolbox for high-dimensional survival analysis. Currently, the package is focused exclusively on knowledge distillation for sparse survival analysis, sometimes also called preconditoning [2, 3]. In the future, we plan to also extend sparsesurv to other techniques useful for (high-dimensional) survival analysis that are not commonly available in Python.
The easiest way to install sparsesurv is currently via PyPi:
pip install sparsesurv
If you want to install directly from Github, you can also install by cloning the repo, or directly piping the repo to pip:
git clone https://github.com/BoevaLab/sparsesurv/
cd sparsesurv
pip install .
pip install git+https://github.com/BoevaLab/sparsesurv.git
If there is sufficient interest, we may also provide a conda package in the future.
If you have a bug report to make or a feature request for something you would like included in sparsesurv in the future, please open a Github issue.
If you have general questions, meaning you are unsure about the usage of sparsesurv, or have other questions about the package that do not seem like a bug or feature request, please use Github discussions.
Documentation and user guides are available on Github pages.
We always welcome new contributors to sparsesurv. If you're interested in contributing, get in touch with us (see Contact) or have a look at the open issues.
If you use any or part of this package, please cite our work. [TODO - add bibtext]
[1] David Wissel, Nikita Janakarajan, Daniel Rowson, Julius Schulte, Xintian Yuan, Valentina Boeva. "sparsesurv: Sparse survival models via knowledge distillation." (2023, under review).
[2] Paul, Debashis, et al. "“Preconditioning” for feature selection and regression in high-dimensional problems." (2008): 1595-1618.
[3] Pavone, Federico, et al. "Using reference models in variable selection." Computational Statistics 38.1 (2023): 349-371.