Skip to contents

Number of Components Needed to Reach a Cumulative Variance Threshold

Source: R/pca_equation.R
pca_component_cutoff.Rd

Number of Components Needed to Reach a Cumulative Variance Threshold

Usage

pca_component_cutoff(pca_mod, threshold = 0.8)

Arguments

pca_mod

A [stats::prcomp()] object or a [run_pca()] result.

threshold

Cumulative share of the total variance, in (0, 1] (default 0.80).

Value

The smallest number of components whose cumulative variance share is at least `threshold`, as an integer with the attribute `cumulative` (the shares for all components). For `prcomp(..., rank. = k)` the shares are of the full variance; a threshold that the retained components cannot reach is an error.

See also

[pca_variance_pct()], [pca_equation_latex()]

Examples

pca_component_cutoff(prcomp(mtcars, scale. = TRUE), 0.9)
#> [1] 4
#> attr(,"cumulative")
#>  [1] 0.6007637 0.8417153 0.8987332 0.9232421 0.9435558 0.9627918 0.9750884
#>  [8] 0.9862612 0.9932655 0.9979960 1.0000000