update_em2: Estimation of the D-DFM model using the EM algorithm

In juhokalle/rmfd4dfm: Estimation of DFMs using RMFD-E parametrization

Estimation of the D-DFM model using the EM algorithm

Description

This function carries out the EM estimation of RMFD model by iterating between E and M step (smoothed_moments8 and max_step, resp.) until convergence

Usage

update_em2(
  params_deep_init,
  sigma_init,
  data_wide,
  tmpl_ss,
  MAXIT = 100,
  VERBOSE = FALSE,
  conv_crit = 0.01
)

Arguments

params_deep_init	initial system and idiosyncratic variance parameter values
sigma_init	initial value of the dim_in x dim_in innovation covariance matrix
data_wide	n x T matrix containing the data
tmpl_ss	tmpl_rmfd_echelon_ss object specifying the model structure
MAXIT	integer denoting maximum number of iterations
VERBOSE	boolean, print estimation progress?

juhokalle/rmfd4dfm documentation built on July 18, 2024, 10:19 p.m.