f1: Estimating F1 score

In loublanpain/BMI585FinalProjectLouBlanpain: What the Package Does (One Line, Title Case)

Description

Computes the F1 score of a classifier, by using recorded data vs data predicted by a classifier.

Usage

f1(pred, truth)

Arguments

pred	A logical (T=positive, F=negative) or numeric (1=positive, 0=negative) vector of positives and negatives provided by the classifier.
truth	A logical (T=positive, F=negative) or numeric (1=positive, 0=negative) vector of positives and negatives measured.

Details

Equation used is 2 x (precision x recall) / (precision + recall). Precision is calculated using function PPV, while recall is calculated using function Sensitivity.

Value

A value from 0 (low F1 score) to 1 (high F1 score).

Examples

pred_values=sample(c(0,1),replace=T,size=100)
true_values=sample(c(0,1),replace=T,size=100)
f1(pred_values,true_values)

loublanpain/BMI585FinalProjectLouBlanpain documentation built on Dec. 21, 2021, 11:49 a.m.