README.md

In HGray384/pcaNet: Probabilistic principal components analysis - covariance estimation and network reconstruction

pcaNet

Probabilistic principal components analysis for covariance matrix estimation and network reconstruction R-package (dev version)

Implementation of the algorithms from Gray and Kirk, (manuscript in preparation) for probabilistic principal components analysis (PPCA) focussed on high-dimensional covariance matrix estimation and network reconstruction. PPCA is a model-based approach for dimension reduction, which may be applied to incomplete data and implicitly performs covariance matrix estimation during model fitting. This package implements a number of PPCA algorithms and provides extra functionality to visualise the estimated covariance matrix, efficiently compute its inverse, estimate the underlying network of variables, and then also visualise this network.

See documentation files for information on the usage of PPCA.

This version can be directly installed from the R console using:

# install.packages("devtools")
devtools::install_github("HGray384/pcaNet")

HGray384/pcaNet documentation built on Nov. 14, 2020, 11:11 a.m.