SINGLE Package Vignette

Given multiple longitudinal time series we are often interested in quantifying the relationship between the time series over time. For example, given fMRI data corresponding to various regions of interest in the brain we may be interested in measuring their statistical dependencies over time. These relationships can subsequently be summarised as graphs or networks where each node corresponds to a time series (e.g., a BOLD time series for a region of interest) and edges between nodes (or their absence) provide information regarding the nature of the pairwise relationship. Here we present a short tutorial and introduction to using the R package SINGLE to estimate such dynamic graphs over time from noisy time series data. The SINGLE package provides an implementation of the Smooth Incremental Graphical Lasso Estimation algorithm, full details of which can be found in [Monti et al., 2013].