louis_information: Get the information matrix using the Louis decomposition...

In cthmm: Latent Continuous Time Markov Chains

Description

Get the information matrix using the Louis decomposition method.

Usage

  louis_information(rates.setup, init.setup,
    emission.setup, delta.list, emission.list, rates.list,
    M1.list, M2.list, rate.param.values = NULL,
    init.param.values = NULL, emission.param.values = NULL,
    durations, emission.counts, init.counts, num.subjects,
    transitions)

Arguments

rates.setup	list rate setup object
init.setup	init setup object
emission.setup	emission setup object
init.list	list with individual initial distributions
emission.list	with individual emission distributions
rates.list	with individual rate matrices
M1.list	list with conditional means of sufficient statistics for each subject
M2.list	list with 2nd moments of sufficient statistics for each subject
rate.param.values	values of rate parameters
init.param.values	values of initial distribution parameters
emission.param.values	of emission parameters
durations	array (dim num.states x num.subjects) with the expected durations
emission.counts	array expected nunber of o_ij 's dim (num states x num.obs.states x num.subjects)
num.subjects	number of subjects in the study

Value

list with the following components

out	Information matrix
complete_info	the complete data information
score_sq	Conditional expectation of the square of complete data information
cross_term	Cross term in the conditional expecation of the complete data information

Author(s)

Jane Lange

cthmm documentation built on May 2, 2019, 5:46 p.m.