varcompare: varcompare function

In clmacleod/highlandr: Random Useful Functions

varcompare function

Description

create function to create a matrix that compares all variable types with quazi correlation measures to assess multicollinearity using the following criteria. vartypes must be "dochotomous" - d , "nominal" - n, or "continuous" - c and the permutation determines what is used to compare. the following are the correlation metric and significance types used and the associated permuations. #phi, x2 (dd); #cramer's v, x2 (dn,nd,nn); #point biserial, t (dc,cd); #omega2, anova (nc,cn); #pearson or spearman (depending on normality-shapirowilks), correlation sigs (cc)

Usage

varcompare(d, vars, vartypes, force_pc = FALSE)

Arguments

d	data frame containing all variables to be compared
vars	the names of all the variables in the dataset
vartypes	a vector of types of the variables. types can be 'd', 'n', or 'c' which are dichotomous, nominal, and continuous respectively

Examples

varcompare_function()

clmacleod/highlandr documentation built on Oct. 11, 2023, 8:04 p.m.