methods for fair.model objects {fairml}

Extract information from fair.model objects

Description

Extract various quantities of interest from an object of class fair.model.

Usage

# methods for all fair.model objects.
## S3 method for class 'fair.model'
coef(object, ...)
## S3 method for class 'fair.model'
residuals(object, ...)
## S3 method for class 'fair.model'
fitted(object, type = "response", ...)
## S3 method for class 'fair.model'
sigma(object, ...)
## S3 method for class 'fair.model'
deviance(object, ...)
## S3 method for class 'fair.model'
logLik(object, ...)
## S3 method for class 'fair.model'
nobs(object, ...)
## S3 method for class 'fair.model'
print(x, digits, ...)
## S3 method for class 'fair.model'
summary(object, ...)
## S3 method for class 'fair.model'
all.equal(target, current, ...)
## S3 method for class 'fair.model'
plot(x, support = FALSE, regression = FALSE, ncol = 2, ...)

# predict() methods.
## S3 method for class 'nclm'
predict(object, new.predictors, new.sensitive, type = "response", ...)
## S3 method for class 'zlm'
predict(object, new.predictors, type = "response", ...)
## S3 method for class 'zlrm'
predict(object, new.predictors, type = "response", ...)
## S3 method for class 'frrm'
predict(object, new.predictors, new.sensitive, type = "response", ...)
## S3 method for class 'fgrrm'
predict(object, new.predictors, new.sensitive, type = "response", ...)

Arguments

object, x, target, current

an object of class fair.model or nclm.

type

a character string, the type of fitted value. If "response", fitted() and predict() will return the fitted values (if the response in the model is continuous) or the classification probabilities (if it was discrete). If "class" and object is a classifier, fitted() and predict() will return the class labels as a factor. If "link" and object is a classifier, fitted() and predict() will return the linear component of the fitted or predicted value, on the scale of the link function.

digits

a non-negative integer, the number of significant digits.

new.predictors

a numeric matrix or a data frame containing numeric and factor columns; the predictors for the new observations.

new.sensitive

a numeric matrix or a data frame containing numeric and factor columns; the sensitive attributes for the new observations.

support

a logical value, whether to draw support lines (diagonal of the first quadrant, horizontal line at zero, etc.) in plot().

regression

a logical value, whether to draw the regression line of the observed values on the fitted values from the model in plot().

ncol

a positive integer, the number of columns the plots will be arranged into.

...

additional arguments, currently ignored.

Author(s)

Marco Scutari