samples_LNR {EMC2} | R Documentation |
An emc object of an LNR model of the Forstmann dataset using the first three subjects
Description
An emc object with a limited number of samples and subjects of the Forstmann dataset. The object is a nested list of lenght three, each list containing the MCMC samples of the respective chain. The MCMC samples are stored in the samples element.
Usage
samples_LNR
Format
An emc object. An emc object is a list with a specific structure and elements, as outlined below.
- data
A list of dataframes, one for each subject included
- par_names
A character vector containing the model parameter names
- n_pars
The number of parameters in the model
- n_subjects
The number of unique subject ID's in the data
- subjects
A vector containing the unique subject ID's
- prior
A list that holds the prior for
theta_mu
(the model parameters). Contains the mean (theta_mu_mean
), covariance matrix (theta_mu_var
), degrees of freedom (v
), and scale (A
) and inverse covariance matrix (theta_mu_invar
)- ll_func
The log likelihood function used by pmwg for model estimation
- samples
A list with defined structure containing the samples, see the Samples Element section for more detail
- grouped
Which parameters are grouped across subjects, in this case none
- sampler_nuis
A sampler list for nuisance parameters (in this case there are none), similarly structured to the overall samples list of one of the MCMC chains.
Samples Element
The samples element of a emc object contains the different types of samples
estimated by EMC2. These include the three main types of samples
theta_mu
, theta_var
and alpha
as well as a number of
other items which are detailed here.
- theta_mu
samples used for estimating the model parameters (group level), an array of size (n_pars x n_samples)
- theta_var
samples used for estimating the parameter covariance matrix, an array of size (n_pars x n_pars x n_samples)
- alpha
samples used for estimating the subject random effects, an array of size (n_pars x n_subjects x n_samples)
- stage
A vector containing what PMwG stage each sample was drawn in
- subj_ll
The winning particles log-likelihood for each subject and sample
- a_half
Mixing weights used during the Gibbs step when creating a new sample for the covariance matrix
- last_theta_var_inv
The inverse of the last samples covariance matrix
- idx
The index of the last sample drawn
- epsilon
The scaling parameter of the proposal distributions for each subject array of size (n_subjects x n_samples)
- origin
From which propoosal distribution the accepted samples in the MCMC chain came, an array of size (n_subjects x n_samples)
Source
https://www.pnas.org/doi/10.1073/pnas.0805903105