gen_decomposed_results_multinom: Decompose results of a mulitnomial regression model

In telliott27-research/KOBdecomp: KOB Decomposition

Decompose results of a mulitnomial regression model

Description

Given a properly expressed multinom regression, decompose the results into an explained and unexplained parts.

Usage

gen_decomposed_results_multinom(
  fitted.mdl,
  df.analysis,
  list_of_IVs,
  list_of_grouping_vars,
  values_of_grouping_vars
)

Arguments

fitted.mdl	A multinom regression object from the nnet package
df.analysis	the dataframe used to run the regression model
list_of_IVs	A character vector indicator the predictor variables included in the probit model
list_of_grouping_vars	A character vector of grouping variables included in the probit model
values_of_grouping_vars	A vector containing all possible values of the grouping variable.

Value

A list containing two objects. One, the fitted OLS model based on the estimated log-odds continuous variable. Two, a data frame containing the decomposition comparing every possible combination of values for the grouping variable.

telliott27-research/KOBdecomp documentation built on March 19, 2022, 7:27 a.m.