Bootstrap PCA Results

An S4 class to store the bootstrap of a principal components analysis.

Note

This class inherits from PCA and MultivariateBootstrap.

See also

Other class: BootstrapCA-class, CA-class, MCA-class, MultivariateAnalysis, MultivariateBootstrap, MultivariateResults, MultivariateSummary, PCA-class, PCOA-class

Author

N. Frerebeau

Examples

## Bootstrap on CA
## Data from Lebart et al. 2006, p. 170-172
data("colours")

## Compute correspondence analysis
X <- ca(colours)

## Bootstrap (30 replicates)
Y <- bootstrap(X, n = 30)

if (FALSE) { # \dontrun{
## Get replicated coordinates
get_replications(Y, margin = 1)
get_replications(Y, margin = 2)
} # }

## Plot with ellipses
viz_rows(Y)
viz_tolerance(Y, margin = 1, level = c(0.68, 0.95))


viz_columns(Y)
viz_tolerance(Y, margin = 2, level = c(0.68, 0.95))


## Plot with convex hulls
viz_columns(Y)
viz_hull(Y, margin = 2)


## Bootstrap on PCA
## Compute principal components analysis
data("iris")
X <- pca(iris)
#> 1 qualitative variable was removed: Species.

## Bootstrap (30 replicates)
Y <- bootstrap(X, n = 30)

## Plot with ellipses
viz_variables(Y)
viz_tolerance(Y, margin = 2, level = c(0.68, 0.95))