distributed quantile regression feature screening

This repository contains the code to reproduce the experimental results from the paper "Communication-Efficient Feature Screening for Ultrahigh-dimensional Data under Quantile Regression".

Prerequisites

• python==3.11.8
• numpy==1.26.4
• scipy==1.12.0
• pandas==2.2.0
• numba==0.59.0
• openpyxl==3.1.2

Usage

• simulator.py contains the code to simulate the data.
• fun.py contains all the functions used in the paper.
• conquer.py contains the code to reproduce the initial values.
• screening.py contains the code to implement the proposed method.
• U_fun.py contains the code to implement the U-statistics method.
• main.py contains the code to reproduce the results in the paper.