f_pca: calculate principle components for a dataset

In erblast/oetteR: Collection of personal R functions

Description

This function is an extended wrapper for prcomp(). I takes a data_ls object created by f_clean_data and calculates the contribution of each variable to each principle component in percent.

Usage

f_pca(data_ls, center = T, scale = T, use_boxcox_tansformed_vars = T,
  include_ordered_categoricals = T)

Arguments

data_ls	data_ls object generated by f_clean_data(), or a named list list( data = <dataframe>, numericals = < vector with column names of numerical columns>)
center	boolean, Default: T
scale	boolean, Default: T
use_boxcox_tansformed_vars	boolean, Default: T
include_ordered_categoricals	boolean, Default: T

Details

Blog post explaining how to calculate contributions

Value

a list with the original data complemented with the principle component vector data of each observation and an object returned by prcomp() supplemented with some extra features

data	dataframe
pca	pca object created by prcomp()

added features of pca:

cos2	The squared rotation vectors.A value between 0 and 1 denotes the amount of contribution of a variable to a specific principle component
vae	percent variance explained
contrib_abs_perc	The absolute contribution of one variable to the variance explained by one principle component in percent. The total contibution adds up to the total contibution of the principle componaent in percent.
contrib_abs_perc_reduced	as above but variables contibuting less than 2.5 percent are grouped
threshold_vae_for_pc_perc	principle components that explain less percent variance than this threshold are dropped

See Also

prcomp

Examples

pca_ls = f_clean_data(mtcars) %>%
f_boxcox() %>%
f_pca()

erblast/oetteR documentation built on May 27, 2019, 12:11 p.m.