Summary table for a bootstrapped Joint Graphical Lasso model
Create a table of bootstrapped means and confidence intervals for all edges of a bootstrapped Joint Graphical Lasso model obtained through GroupNetworkBoot .
BootTable(BootOut)
BootOut: The output from GroupNetworkBootSummary table of the output of GroupNetworkBoot
Var1: Nodes included in each edge
Var2: Nodes included in each edge
edges: Edge identifier
sample: sample value of each edge
boot.mean: mean of boostrapped values of each edge
ci.lb: lower bound of the .95 confidence interval
ci.ub: upper bound of the .95 confidence interval
boot.zero: proportion of bootstraps, in which an edge was estimated as equal to zero (i.e., 0= edge not estimated as zero throughout bootstraps; 1= edge estimated as zero in all bootstraps)
boot.pos: Proportion of bootstraps in which an edge was estimated as >0 (i.e., positive)
boot.neg: Proportion of bootstraps in which an edge was estimated as <0 (i.e., negative)
g: group in which the edge was estimated
Epskamp, S., Borsboom, D., & Fried, E. I. (2018). Estimating psychological networks and their accuracy: A tutorial paper. Behavior Research Methods, 50(1), 195–212. https://doi.org/10.3758/s13428-017-0862-1 Danaher, P., Wang, P., & Witten, D. M. (2014). The joint graphical lasso for inverse covariance estimation across multiple classes. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 76(2), 373–397. https://doi.org/10.1111/rssb.12033
Nils Kappelmann n.kappelmann@gmail.com, Giulio Costantini
JGL, qgraph, parcor
Useful links