performance_measures: Performance measures for ROCR::prediction object

In GegznaV/spHelper: Extension of Package `hyperSpec` and Various Convenience Functions

Performance measures for ROCR::prediction object

Description

Function get_performance calculates the best performance measure when prediction object is given.

Usage

get_performance(pred, measure)

get_max(pred, FUN)

AUC(pred)

tp(pred)

fp(pred)

fn(pred)

tn(pred)

cutoff(pred)

Se(pred)

Sp(pred)

J(pred)

Bac(pred)

Acc(pred)

Kappa(pred)

Wkappa(pred)

kappa_helper(pred, FUN_)

make_conf_matrix(TP, FN, FP, TN)

Arguments

pred	An object of class prediction from package ROCR.
measure	(string(1)) A string with the name of classification performance measure to use. Currently available options: "bac" - for balanced accuracy (mean of sensitivity and specificity); "kappa" - for Cohens kappa; "wkappa" - for weighted Cohens kappa; "j" - for Youden's index; "auc" - for area under the ROC curve; "acc" - for accuracy (total proportion of correctly identified cases).
FUN	Function to apply (one of Acc, Bac, J, Kappa, Wkappa, etc.)
FUN_	Function to apply (either measure_kappa or measure_wkappa)
TP	Number of true positives.
FN	Number of false negatives.
FP	Number of false positives.
TN	Number of true negatives.

Value

Function get_performance returns a numeric vector with 2 elements:

• the first element is the highest value of selected performance measure;

• the second element is either corresponding cut-off value, or NA if the measure is "auc".

GegznaV/spHelper documentation built on April 16, 2023, 1:42 p.m.