mctest.ergoInfo: Monte-Carlo Test for the Ergodicity Information Index

Description Usage Arguments Value Author(s) Examples

View source: R/mctest.ergoInfo.R

Description

Computes a Monte-Carlo Test for the Ergodicity Information Index, comparing the empirical Ergodicity Information index to values obtained in a Monte-Carlo simulation in which all individuals have a similar latent structure. The p-values in the Monte-Carlo test can be calculated as (sum(EII>=MC.EII)+1)/(iter+1) and as (sum(EII<=MC.EII)+1)/(iter+1), where EII is the empirical Ergodicity Information Index, MC.EII is the values of the Ergodicity Information Index obtained in the simulation, and iter is the number of random samples generated in the simulation. The two-sided p-value is computed as two times the lowest p-value. In the Monte-Carlo Test for the Ergodicity Information Index, the null hypothesis is that the empirical value of EII is equal to the Monte-Carlo value of EII obtained in multiple individuals with a similar latent structure. Small values of p indicate that is very unlikely to obtain an EII as large as the one obtained in the empirical sample if the null hypothesis is true, thus there is convincing evidence that the empirical Ergodicity Information Index is different than it could be expected if all individuals had a similar latent structure, conditioned on the parameters used to simulate the data.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
mctest.ergoInfo(
  iter,
  N,
  EII,
  use,
  variab,
  timep,
  nfact,
  error,
  dfm,
  loadings,
  autoreg,
  crossreg,
  var.shock,
  cov.shock,
  embed,
  tau,
  delta,
  derivatives,
  model,
  model.args = list(),
  algorithm = c("walktrap", "louvain"),
  algorithm.args = list(),
  corr,
  ncores,
  ...
)

Arguments

iter

Numeric integer. Number of random samples to generate in the Monte-Carlo simulation. At least 500 is recommended

N

Numeric integer. Number of individuals to simulate data from, using the simDFM function.

EII

Numeric. Empirical Ergodicity Information Index obtained via the ergoInfo function.

use

Character. A string indicating what network element will be used to compute the algorithm complexity in the ergoInfo function, the list of edges or the weights of the network. Defaults to use = "edge.list". Current options are:

  • edge.list Calculates the algorithm complexity using the list of edges.

  • weights Calculates the algorithm complexity using the weights of the network.

variab

Number of variables per factor.

timep

Number of time points.

nfact

Number of factors.

error

Value to be used to construct a diagonal matrix Q. This matrix is p x p covariance matrix Q that will generate random errors following a multivariate normal distribution with mean zeros. The value provided is squared before constructing Q.

dfm

A string indicating the dynamical factor model to use. Defaults to "DAFS". Current options are:

  • DAFS Simulates data using the direct autoregressive factor score model. This is the default method

  • RandomWalk Simulates data using a dynamic factor model with random walk factor scores.

loadings

Magnitude of the loadings.

autoreg

Magnitude of the autoregression coefficients. Default is "autoreg = 0.8".

crossreg

Magnitude of the cross-regression coefficients. Default is "crossreg = 0.1".

var.shock

Magnitude of the random shock variance. Default is "var.shock = 0.18".

cov.shock

Magnitude of the random shock covariance Default is "cov.shock = 0.36".

embed

Integer. Number of embedded dimensions (the number of observations to be used in the Embed function). For example, an "embed = 5" will use five observations to estimate a single derivative. Defaults to embed = 5.

tau

Integer. Number of observations to offset successive embeddings in the Embed function. A tau of one uses adjacent observations. Default is "tau = 1".

delta

Integer. The time between successive observations in the time series. Default is "delta = 1".

derivatives

Integer. The order of the derivative to be used in the EGA procedure. Default to 1.

model

Character. A string indicating the method to use. Defaults to glasso. Current options are:

  • glasso Estimates the Gaussian graphical model using graphical LASSO with extended Bayesian information criterion to select optimal regularization parameter. This is the default method

  • TMFG Estimates a Triangulated Maximally Filtered Graph

model.args

List. A list of additional arguments for EBICglasso.qgraph or TMFG

algorithm

A string indicating the algorithm to use or a function from igraph

Current options are:

  • walktrap Computes the Walktrap algorithm using cluster_walktrap

  • louvain Computes the Walktrap algorithm using cluster_louvain

algorithm.args

List. A list of additional arguments for cluster_walktrap, cluster_louvain, or some other community detection algorithm function (see examples)

corr

Type of correlation matrix to compute. The default uses cor_auto. Current options are:

  • cor_auto Computes the correlation matrix using the cor_auto function from qgraph.

  • pearson Computes Pearson's correlation coefficient using the pairwise complete observations via the cor function.

  • spearman Computes Spearman's correlation coefficient using the pairwise complete observations via the cor function.

ncores

Numeric. Number of cores to use in computing results. Defaults to parallel::detectCores() / 2 or half of your computer's processing power. Set to 1 to not use parallel computing. Recommended to use maximum number of cores minus one

If you're unsure how many cores your computer has, then use the following code: parallel::detectCores()

...

Additional arguments. Used for deprecated arguments from previous versions of EGA

Value

Returns a list containing:

mc.ergoInfo

The values of the Ergodicity Information Index obtained in the Monte-Carlo Simulation

p.value.twosided

The p-value of the Monte-Carlo test for the Ergodicity Information Index. The null hypothesis is that the empirical Ergodicity Information index is equal to the expected value of the EII if the all individuals had similar latent structures.

effect

Indicates wheter the empirical EII is greater or less then the Monte-Carlo obtained EII.

plot.dist

Histogram of the bootstrapped ergodicity information index

Author(s)

Hudson Golino <hfg9s at virginia.edu>

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
## Not run: 
\donttest{
dyn1 <- dynEGA.ind.pop(data = sim.dynEGA, n.embed = 5, tau = 1,
                      delta = 1, id = 21, group = 22, use.derivatives = 1,
                    model = "glasso", ncores = 2, corr = "pearson")

eii1 <- ergoInfo(data = dyn1)$EII

dist.ergoinfo <- mctest.ergoInfo(iter = 10, N = 10, EII = eii1,
variab = 4,
timep = 100, nfact = 2, error = 0.05, dfm = "DAFS", loadings = 0.55, autoreg = 0.8,
crossreg = 0.1, var.shock = 0.18, cov.shock = 0.36, embed = 5, tau=1, delta=1, derivatives=1,
model = "glasso", ncores = 2, corr = "pearson")
}
## End(Not run)

EGAnet documentation built on Feb. 17, 2021, 1:06 a.m.