withinCov: Within-class Covariance Matrix

In DiscriMiner: Tools of the Trade for Discriminant Analysis

Description

Calculates the within-class covariance matrix

Usage

  withinCov(variables, group, div_by_n = FALSE)

Arguments

variables	matrix or data frame with explanatory variables (No missing values are allowed)
group	vector or factor with group memberships (No missing values are allowed)
div_by_n	logical indicating division by number of observations

Details

When div_by_n=TRUE the covariance matrices are divided by n (number of observations), otherwise they are divided by n-1

Author(s)

Gaston Sanchez

See Also

withinSS, betweenCov, totalCov

Examples

## Not run: 
  # load iris dataset
  data(iris)

  # within-class covariance matrix (dividing by n-1)
  withinCov(iris[,1:4], iris[,5])

  # within-class covariance matrix (dividing by n)
  withinCov(iris[,1:4], iris[,5], div_by_n=TRUE)
  
## End(Not run)

Example output

             Sepal.Length Sepal.Width Petal.Length Petal.Width
Sepal.Length   0.26145101  0.09147651   0.16526577  0.03788591
Sepal.Width    0.09147651  0.11383893   0.05450201  0.03227114
Petal.Length   0.16526577  0.05450201   0.18270201  0.04209262
Petal.Width    0.03788591  0.03227114   0.04209262  0.04131946
             Sepal.Length Sepal.Width Petal.Length Petal.Width
Sepal.Length   0.25970800  0.09086667   0.16416400  0.03763333
Sepal.Width    0.09086667  0.11308000   0.05413867  0.03205600
Petal.Length   0.16416400  0.05413867   0.18148400  0.04181200
Petal.Width    0.03763333  0.03205600   0.04181200  0.04104400

DiscriMiner documentation built on May 1, 2019, 10:32 p.m.