Extract edge selection frequency in bootstrap subsamples

Extracts edge selection frequency in networks reconstructed from bootstrap subsamples during the stars stability selection procedure, as either a matrix or a named vector. In the latter case, edge names follow igraph naming convention.

extract_probs(
  Robject,
  penalty = NULL,
  index = NULL,
  crit = c("StARS", "BIC", "EBIC"),
  format = c("matrix", "vector"),
  tol = 1e-05
)

Arguments

Robject

an object with class PLNnetworkfamily, i.e. an output from PLNnetwork()

penalty

penalty used for the bootstrap subsamples

index

Integer index of the model to be returned. Only the first value is taken into account.

crit

a character for the criterion used to performed the selection. Either "BIC", "ICL", "EBIC", "StARS", "R_squared". Default is ICL for PLNPCA, and BIC for PLNnetwork. If StARS (Stability Approach to Regularization Selection) is chosen and stability selection was not yet performed, the function will call the method stability_selection() with default argument.

format

output format. Either a matrix (default) or a named vector.

tol

tolerance for rounding error when comparing penalties.

Value

Either a matrix or named vector of edge-wise probabilities. In the latter case, edge names follow igraph convention.

Examples

data(trichoptera)
trichoptera <- prepare_data(trichoptera$Abundance, trichoptera$Covariate)
nets <- PLNnetwork(Abundance ~ 1 + offset(log(Offset)), data = trichoptera)
#> 
#>  Initialization...
#>  Adjusting 30 PLN with sparse inverse covariance estimation
#> 	Joint optimization alternating gradient descent and graphical-lasso
#> 	sparsifying penalty = 3.65921 
	sparsifying penalty = 3.379906 
	sparsifying penalty = 3.121921 
	sparsifying penalty = 2.883627 
	sparsifying penalty = 2.663522 
	sparsifying penalty = 2.460218 
	sparsifying penalty = 2.272432 
	sparsifying penalty = 2.098979 
	sparsifying penalty = 1.938766 
	sparsifying penalty = 1.790781 
	sparsifying penalty = 1.654092 
	sparsifying penalty = 1.527837 
	sparsifying penalty = 1.411218 
	sparsifying penalty = 1.303501 
	sparsifying penalty = 1.204006 
	sparsifying penalty = 1.112105 
	sparsifying penalty = 1.027219 
	sparsifying penalty = 0.9488126 
	sparsifying penalty = 0.8763906 
	sparsifying penalty = 0.8094964 
	sparsifying penalty = 0.7477083 
	sparsifying penalty = 0.6906363 
	sparsifying penalty = 0.6379207 
	sparsifying penalty = 0.5892287 
	sparsifying penalty = 0.5442534 
	sparsifying penalty = 0.502711 
	sparsifying penalty = 0.4643395 
	sparsifying penalty = 0.4288969 
	sparsifying penalty = 0.3961595 
	sparsifying penalty = 0.365921 

#>  Post-treatments
#>  DONE!
if (FALSE) {
stability_selection(nets)
probs <- extract_probs(nets, crit = "StARS", format = "vector")
probs
}

if (FALSE) {
## Add edge attributes to graph using igraph
net_stars <- getBestModel(nets, "StARS")
g <- plot(net_stars, type = "partial_cor", plot=F)
library(igraph)
E(g)$prob <- probs[as_ids(E(g))]
g
}