pcabootAG: Bootstrap Analysis of PCA

In s-huebler/rotateS21: Shiny Data Rotation

Description

This function allows the user to make estimates for the eigenvalues and eigenvectors of a covariance or correlation matrix. The function provides bootstrap and Anderson-Girshick estimates for the lambdas (eigenvalues). The function provides bootstrap estimates for the eigenvectors. The iter argument controls the number of times the bootstrap is run, and the alpha determines the confidence level printed to the console. Histograms for the lambdas and violin plots for the eigenvectors are printed as well. The three types of confidence intervals are stored as lists to be called if the results of the function are stored as an object.

Usage

pcabootAG(df, alpha = 0.05, iter, covar = TRUE)

Arguments

df	, df is a data frame of quantitative variables
alpha,	alpha is a confidence level
iter,	iter is the number of desired bootstrap iterations
covar,	covar is a logical arg where TRUE uses covariance and FALSE uses correlation

Value

A list a confidence intervals

Examples

df <- iris[c(2,3)]
pcabootAG(df, 0.01, iter=1000, covar=TRUE )

s-huebler/rotateS21 documentation built on Dec. 22, 2021, 8:21 p.m.