Selects the best set of predictors among the given sets by means of AIC.

Computes all the models defined by a list of groups of predictors Returns the best model according to AIC and id of the corresponding predictors in the list The different models are computed in parallel

nrmChoose(
  adj,
  w.list,
  xi = NULL,
  directed,
  selfloops,
  pval = 0.05,
  init = NULL,
  ncores = NULL
)

nrm_choose(
  adj,
  w.list,
  xi = NULL,
  directed,
  selfloops,
  pval = 0.05,
  init = NULL,
  ncores = NULL
)

Arguments

adj

adjacency matrix

w.list

nrmPredictor object. Nested list of predictors to be selected.

xi

Xi matrix (optional). defaults to scm Xi matrix.

directed

logical. Is the network directed?

selfloops

logical. Does the network contain selfloops?

pval

numeric. the significance at which computing confidence intervals. defaults to 0.05

init

initial values for the MLE numerical maximisation. (See nrm.)

ncores

Number of cores for parallelisation of selection process. (optional) Defaults to number of available cores - 1.

Value

list containing the best model according to AIC and id of the corresponding predictors in the list