Boot4Mean: Boot4Mean computes a brick of bootstrap estimates for the...

In HerveAbdi/PTCA4CATA: Partial Triadic Analysis (PTCA) for Check All That Apply (CATA) Data

Boot4Mean computes a brick of bootstrap estimates for the means of groups of observations described by several variables (e.g., factor scores).

Description

Boot4Mean computes a brick of bootstrap estimates for the means of groups of observations described by several variables, as for example, for groups of observations described by factors scores computed from a Principal Component Analysis or or a Multiple Correspondence Analysis. The groups of observations are identified by a design factor. Observations are Bootstrapped within their groups (so each Bootstrap sample has the same N per group as the original groups).

Usage

Boot4Mean(Data, design, niter = 100, suppressProgressBar = TRUE)

Arguments

Data	An observations by variables data set.
design	A vector giving the design for the observations: observations with the same number belong to the same group
niter	(Default = 100) the number of bootstrap iterations.
suppressProgressBar	(default = TRUE), when TRUE suppress the progress bar.

Details

Use with PCA-like techniques as a step before plotting confidence interval ellipses for group means

Value

A list with BootCube: a groups by variables by iterations brick of bootstrapped means, and GroupMeans: the groups by variables original means of the groups and BootsrappedGroupMeans: the groups by variables bootstrapped means of BootCube.

Author(s)

Herve Abdi

Examples

## Not run: 
if(interactive()){
 #EXAMPLE1
 }

## End(Not run)

HerveAbdi/PTCA4CATA documentation built on July 17, 2022, 5:41 a.m.