/fssemR

An optimizer of Fused-Sparse Structural Equation Models, which is the state-of-the-art (sota) jointly fused sparse maximum likelihood function for structural equation models proposed by Xin Zhou and Xiaodong Cai

Primary LanguageR

fssemR

CRAN status CRAN download

fssemR is a package that ultilizes the Proximal Alternating Linearized Maximal to solve the non-convex non-smooth jointly fused sparse structrual equation model.

Installation

fssemR package contains a lot of necessary scripts to analyze large dataset such as microarray and SNP data from GEO database, so it has not been submitted to CRAN yet for these non-standard directory. To install fssemR, you need a C++ compiler such as g++ or clang++ with C++11 feature, and for Windows users, the Rtools software is needed (unless you can configure the toolchain by yourself).

The installation follows the typical way of R packages on Github:

library(devtools)
install_github("Ivis4ml/fssemR")

Now, fssemR package is uploaded on CRAN. So you will install it via CRAN

install.packages("fssemR")

Vignette

fssemR-introduction

Citation

Xin Zhou, Xiaodong Cai, Inference of differential gene regulatory networks based on gene expression and genetic perturbation data, Bioinformatics, , btz529, https://doi.org/10.1093/bioinformatics/btz529

Update

Upgrade to Eigen 0.3.4 compatible

Acknowledgement

Thank should go to Yilun Zhang 3100105044@zju.edu.cn/zhangyilun.123@bytedance.com for his contribution to the optimization for this package in math and implementation.