In git-tjj/DtracePLN: DtracePLN: Sparse precision matrix estimation via lasso penalized D-trace loss in the multivariate Poisson log-normal model

knitr::opts_chunk$set(echo = TRUE)

DtracePLN: Dtrace method for poisson log-normal model

DtracePLN is a package for calculating the precision matrix of poission log-normal model using moment method and dtrace loss. We first calculate the covariance matrix by moment method and then calculate the precision matrix by dtrace loss using R package EQUAL.

Installation

DtracePLN is available on Github.

R Package installation

• Before installing DtracePLN package, please install cescwang85-EQUAL-da9cbbb.tar.gz on Github.
• For the development version, use the github install

remotes::install_github("https://github.com/git-tjj/DtracePLN")

Usage and main functions

The package comes with a single-cell RNA sequence data about Interferon β-Stimulated PBMCs to present the functionality of main function.

DtracePLN

library(DtracePLN)
## simulation data for testing
data("obs_mat")
DtracePLN_res_sim<-DtracePLN::DtracePLN(obs_mat = DtracePLN::obs_mat,
Sd_est = "GMPR",lambda_vec = NULL , n_lambda = 50 , lambda_valuemax0 = 10)

## real data analysis
data("realdatamatrix")
DtracePLN_res_real<-DtracePLN::DtracePLN(obs_mat = DtracePLN::realdatamatrix,
Sd_est = "GMPR",lambda_vec = NULL , n_lambda = 50 , lambda_valuemax0 = 10)

data_generator

library(MASS)
data_generator_res<-DtracePLN::data_generator(n = 100, p = 50, densy_degree = 0.05,
sd_ls = 0.1, mean_ls = log(10),value_nondiagonal = 0.3)

Reference

Please cite our work using the following references:

git-tjj/DtracePLN documentation built on Jan. 30, 2023, 1:47 a.m.