ParetoElnet: ParetoElnet

In Diversity-ParetoOptimal/ParetoR: Estimates Pareto-Optimal solution for diversity hiring

ParetoElnet

Description

Command function to run the regularized tradeoff curve algorithm.

Usage

ParetoElnet(
  prop,
  sr,
  d,
  R,
  Spac = 10,
  lambda.param = 0,
  alpha.param = 1,
  graph = TRUE,
  display_solution = TRUE
)

Arguments

prop	Proportion of minority applicants in full applicant pool
sr	Selection ratio
d	Subgroup difference
R	Correlation matrix
Spac	Number of Pareto points; default is 10
lambda.param	lambda parameter, adjusts the magnitude of regularization within the elastic net function; default value is 0
alpha.param	alpha parameter, adjusts the extent of ridge and LASSO within the elastic net function; default value is 1
graph	If TRUE, plots will be generated for Pareto-optimal curve and predictor weights
display_solution	If TRUE, Pareto-optimal solutions will be printed

Value

A list containing the criterion values (Pareto_Fmat) and predictor weights (Pareto_Xmat) of Pareto-optimal solutions

Examples

# Specify inputs
# (1) Proportion of minority applicants (prop) = (# of minority applicants)/(total # of applicants)
prop <- 1/4
# (2) Selection ratio (sr) = (# of selected applicants)/(total # of applicants)
sr <- 0.10
# (3) Subgroup differences (d): standardized mean differences between minority
# and majority subgroups (i.e., majority - minority), on each predictor (in applicant pool)
d <- c(1.00, 0.23, 0.09, 0.33)
# (4) Correlation matrix (R) = criterion & predictor inter-correlation matrix (in applicant pool)
# Format: Predictor_1, ..., Predictor_n, Criterion
R <- matrix(c(1, .24, .00, .19, .30,
              .24, 1, .12, .16, .30,
              .00, .12, 1, .51, .18,
              .19, .16, .51, 1, .28,
              .30, .30, .18, .28, 1),
            (length(d)+1),(length(d)+1))
# Fit Regularized Pareto-optimal model
out = ParetoElnet(prop, sr, d, R)

Diversity-ParetoOptimal/ParetoR documentation built on Feb. 9, 2024, 1:06 a.m.