This repository contains some main numerical experiments of synthetic datasets and real-world dataset described in the paper "Simplex Constrained Sparse Optimization via Tail Screening".

tailscreening.py: implementation of our proposed PERMITS method
models.py: implementations of some comparative methods
utils.py: provides some utility functions

simulation_tol/simu_tol: code for simulation with different tolerance params
simulation_tol/plot_acc_error: plot function for Accuracy and Error
simulation_tol/plot_time: plot function for Time

Comparing experiments of our method PERMITS (with different tolerance parameters) with some SOTA methods.

from tailscreening import TailScreening

## generate samples
X, y = ...

# fitting data
model = TailScreening()
model.fit(X=X, y=y)
w_est = model.coef_

Please cite the following publications if you make use of the material here.

Peng Chen, Jin Zhu, Junxian Zhu, & Xueqin Wang (2025). Simplex Constrained Sparse Optimization via Tail Screening. Journal of Machine Learning Research, 26(159), 1–38.

The corresponding BibteX entry is:

@article{JMLR:v26:24-0010,
  author  = {Peng Chen and Jin Zhu and Junxian Zhu and Xueqin Wang},
  title   = {Simplex Constrained Sparse Optimization via Tail Screening},
  journal = {Journal of Machine Learning Research},
  year    = {2025},
  volume  = {26},
  number  = {159},
  pages   = {1--38},
  url     = {http://jmlr.org/papers/v26/24-0010.html}
}