ci_test: Test for conditional independence

In lg: Locally Gaussian Distributions: Estimation and Methods

Description

Perform a test for conditional independence between the first two variables in the data set, given the remaining variables.

Usage

ci_test(lg_object, h = function(x) x^2, S = function(y) rep(T,
  nrow(y)), n_rep = 500, nodes = 100, M = NULL, M_sim = 1500,
  M_corr = 1.5, n_corr = 1.2, extend = 0.3, return_time = TRUE)

Arguments

lg_object	An object of type lg, as produced by the lg_main-function
h	The h-function used in the calculation of the test statistic. The default value is h(x) = x^2.
S	The integration area in the test statistic. Logical function that takes grid points as argument.
n_rep	The number of replicated bootstrap samples
nodes	Either the number of equidistant nodes to generate, or a vector of nodes supplied by the user
M	The value for M in the accept-reject algorithm if already known
M_sim	The number of replicates to simulate in order to find a value for M
M_corr	Correction factor for M, to be on the safe side
n_corr	Correction factor for n_new, so that we mostly will generate enough observations in the first go
extend	How far to extend the grid beyond the extreme data points when interpolating, in share of the range
return_time	Measure how long the test takes to run, and return along with the test result

lg documentation built on Dec. 5, 2019, 5:13 p.m.