roc_selection: Calcuations useful for determining the best number of bins...

In tmc2737/t2roc: Type-2 ROC

Description

This function takes a dataframe and performs a series of calcuations in a step-wise fashion according to a maximum, minimum, and step numbers of bins. The "optimal" number of bins for a type-2 ROC calcuation.

Usage

roc_selection(data, varname, outcomevar, low_bin, high_bin, step_bin)

Arguments

data	a dataframe that holds the judgment and outcome variables.
varname	the name of the variable that is to be binned.
outcomevar	the name of the outcome variable that the hit and false alarm bin calculations are based on.
low_bin	a positive integer indicating the lowest number of bins for the calculations.
high_bin	a positive integer indicating the highest number of bins for the calculations.
step_bin	a positive integer indicating the step-wise progression through bin numbers.

Examples

# Randomly select 100 integers from 0 - 100.
set.seed(2)
a <- sample(0:100,100,replace=TRUE)

# Randomly select 100 integers of 0 or 1.
set.seed(2)
b <- sample(0:1,100,replace=TRUE)

# Send to data frame.
c <- data.frame(a,b)

# Send to function; output bins + calculations
roc_selection(c, varname = "a", outcomevar = "b", 2, 10, 1)

tmc2737/t2roc documentation built on May 29, 2019, 9:31 a.m.