Skip to contents

PCA

Source: R/pca-wrapper.R
PCA.Rd

Principal Component Analysis (PCA) for dimensionality reduction and analysis of multivariate acoustic data.

Arguments

.xptr

Not for direct use. External pointer to the underlying C++ PCA object; set internally when a method returns a new PCA (for example pca_from_matrix).

Value

A PCA object with methods for querying components, eigenvalues, and projections.

Details

PCA is commonly used in phonetics for vowel space analysis, speaker normalization, acoustic feature extraction, and data visualization.

See also

Discriminant

Examples

set.seed(1)
data <- cbind(f1 = rnorm(20, 500, 50), f2 = rnorm(20, 1500, 100))
pca <- pca_from_matrix(data)
pca$get_number_of_components()
#> [1] 2
pca$get_eigenvalues()
#> [1] 7725.581 1952.273