pcabootAG: PCA and Bootstrapping

In leahpom/MATH5793POMERANTZ: Math 5793 R functions

Description

This is a function for the PCA bootstrapping assignment. It performs bootstrapping on the eigenvalues and eigenvectors of either the sample covariance (S) or the sample correlation (R), then it outputs plots of the bootstrapped estimates, as well as confidence intervals. It also outputs Anderson and Girschick confidence intervals for the λ values

Usage

pcabootAG(data, alpha, iter, cov = TRUE)

Arguments

data	a data frame
alpha	the significance level
iter	the number of iterations to use when bootstrapping
cov	a boolean for if we want to use S (cov = TRUE) or R (cov = FALSE) for the analysis

Value

a barplot for the bootstrapped values of each λ[i]

a violin plot for the bootstrapped values for each e_ik, i,k = 1,2,...p, where p is the number of variables

a list containing confidence intervals for each λ[i] and e_ik, i,k = 1,2,...p from bootstrapping, and confidence intervals for each λ[i] using the Anderson and Gerschick method

Examples

data <- as.data.frame(datasets::mtcars)
pcabootAG(data[,1:3], alpha = 0.05, iter = 100)

leahpom/MATH5793POMERANTZ documentation built on May 10, 2021, 9:52 a.m.