Estimates statistical significance of association between variables and their latent variables, from a parametric jackstraw procedure.
jackstraw.parametric(dat, FUN = function(x) fast.svd(x)$v[, 1:r, drop =
FALSE], noise = function(x) rnorm(x, mean = 0, sd = 1), r = NULL,
r1 = NULL, s = NULL, B = NULL, covariate = NULL, verbose = TRUE,
seed = NULL)a data matrix with m rows as variables and n columns as observations.
provide a function to estimate LVs. Must output r estimated LVs in a n*r matrix.
specify a parametric distribution to generate a noise term.
a number of significant latent variables.
a numeric vector of latent variables of interest.
a number of ``synthetic'' null variables. Out of m variables, s variables are independently permuted.
a number of resampling iterations.
a model matrix of covariates with n observations. Must include an intercept in the first column.
a logical indicator as to whether to print the progress.
a seed for the random number generator.
jackstraw.parametric returns a list consisting of
the m p-values of association tests between variables and their principal components
the observed F-test statistics
the s*B null F-test statistics
This function estimates statistical significance of association between variables and latent variables
using a parametric distribution of a noise term. A small number s of observed variables are replaced by
synthetic null variables generated from a specified distribution (such as Normal(0,1)).
After applying a latent variable estimation function on this newly generated matrix (with s synthetic nulls
and m-s intact observed variables), F-test statistics between estimated latent variables and s synthetic nulls
are called the jackstraw statistics. P-values are computed by comparing observed F-test statistics against s*B jackstraw statistics.
Note that unless you have a strong reason to use a parametric distribution, it is advised to use the non-parametric jackstraw.