plot_fairmodels: Plot fairmodels

In fairmodels: Flexible Tool for Bias Detection, Visualization, and Mitigation

Plot fairmodels

Description

Easier access to all plots in fairmodels. Provide plot type (that matches to function name), pass additional parameters and plot.

Usage

plot_fairmodels(x, type, ...)

## S3 method for class 'explainer'
plot_fairmodels(x, type = "fairness_check", ..., protected, privileged)

## S3 method for class 'fairness_object'
plot_fairmodels(x, type = "fairness_check", ...)

## Default S3 method:
plot_fairmodels(x, type = "fairness_check", ...)

Arguments

x	object created with fairness_check or with explain
type	character, type of plot. Should match function name in fairmodels. Default is fairness_check.
...	other parameters passed to fairmodels functions.
protected	factor, vector containing sensitive attributes such as gender, race, etc...
privileged	character/factor, level in factor denoting privileged subgroup

Details

types (function names) available:

• fairness_check

• stack_metrics

• fairness_heatmap

• fairness_pca

• fairness_radar

• group_metric

• choose_metric

• metric_scores

• performance_and_fairness

• all_cutoffs

• ceteris_paribus_cutoff

Value

ggplot2 object

Examples

data("german")

y_numeric <- as.numeric(german$Risk) - 1

lm_model <- glm(Risk ~ .,
  data = german,
  family = binomial(link = "logit")
)
explainer_lm <- DALEX::explain(lm_model, data = german[, -1], y = y_numeric)

# works with explainer when protected and privileged are passed
plot_fairmodels(explainer_lm,
  type = "fairness_radar",
  protected = german$Sex,
  privileged = "male"
)

# or with fairness_object
fobject <- fairness_check(explainer_lm,
  protected = german$Sex,
  privileged = "male"
)

plot_fairmodels(fobject, type = "fairness_radar")

fairmodels documentation built on Aug. 24, 2022, 1:05 a.m.