msu (version 0.0.1)

total_correlation: Estimation of total correlation for a set of categorical random variables.

Description

Total Correlation is a generalization of information gain (IG) to measure the dependency of a set of categorical random variables (see https://en.wikipedia.org/wiki/Total_correlation).

Usage

total_correlation(table_variables, table_class)
C(table_variables, table_class)

Arguments

table_variables

A list of factors as categorical variables.

table_class

A factor representing the class of the case.

Value

Total correlation estimation for the variable set table.variables, table.class.

Examples

# NOT RUN {
total_correlation(list(factor(c(0,1)), factor(c(1,0))), factor(c(0,0)))
total_correlation(list(factor(c('a','b')), factor(c('a','b'))),
    factor(c('a','b')))
# }
# NOT RUN {
total_correlation(list(factor(c(0,1)), factor(c(1,0))), c(0,0))
total_correlation(c(factor(c(0,1)), factor(c(1,0))), c(0,0))
# }