Horn's parallel analysis – suggested number of factors.

Native Horn's parallel analysis: observed eigenvalues vs the 95th percentile of random-data eigenvalues.

Usage

morie_psymet_parallel(data, nsim = 100, seed = 42)

Arguments

data

Numeric matrix or data.frame of items.

nsim

Integer; number of simulated random datasets (default 100).

seed

Integer; RNG seed for reproducibility.

Value

An integer scalar: the number of components to retain by parallel analysis.

Examples

set.seed(1)
f <- rnorm(60)
X <- sapply(1:5, function(j) sqrt(0.6) * f + sqrt(0.4) * rnorm(60))
colnames(X) <- paste0("i", 1:5)
k <- morie_psymet_parallel(X, nsim = 20, seed = 1)
k
#> [1] 1