ClassMetrics: Get classification accuracy metrics

In Spiritspeak/skMisc: Sercan Kahveci's Miscellaneous Functions

Get classification accuracy metrics

Description

Get classification accuracy metrics

Usage

ClassMetrics(x, y)

Arguments

x	The true class memberships
y	The predicted class memberships

Value

A vector with the following accuracy metrics:

• Accuracy: acc

• Chance accuracy: chance_acc

• Cohen's Kappa: kappa

• Positive predictive value: ppv

• Chance positive predictive value: chance_ppv

• Kappa of the positive predictive value: kappa_ppv

• Negative predictive value: npv

• Chance negative predictive value: chance_npv

• Kappa of the negative predictive value: kappa_npv

• Sensitivity: sens

• Specificity: spec

Author(s)

Sercan Kahveci

Examples

mymod <- glm(am ~ cyl + disp + hp,
             family="binomial", 
             data = mtcars)

newpred <- as.numeric(predict(mymod, 
                              data=mtcars,
                              type="response") > 0.5)

ClassMetrics(x=mtcars$am, y=newpred)

Spiritspeak/skMisc documentation built on April 12, 2025, 5:40 a.m.