LDparam set — LD_param_est • exametrika

A function that extracts only the estimation of graph parameters after the rank estimation is completed.

Usage

LD_param_est(
  tmp,
  adj_list,
  classRefMat,
  ncls,
  smoothpost,
  beta1 = 2,
  beta2 = 2,
  fit_only = FALSE,
  bench = NULL
)

tmp: tmp
adj_list: adj_list
classRefMat: values returned from emclus
ncls: ncls
smoothpost: smoothpost
beta1: Beta distribution parameter 1 for prior density. Default is 2.
beta2: Beta distribution parameter 2 for prior density. Default is 2. Unlike the other network models (which default to 1), the default of 2 follows the original Mathematica implementation of LDLRA.
fit_only: If TRUE, return only the BIC (list with element BIC), skipping the posterior distribution, IRP, and full TestFit construction. Used by the LDLRA_PBIL fitness loop, where only the BIC is consumed. Requires bench.
bench: Cached benchmark-model statistics from BNM_bench_stats(), consumed only when fit_only = TRUE. They depend only on the data, so the structure-search loop computes them once instead of re-running TestFit() per candidate.