suggest_probCut: Compare Class Probability Distribution based on Probability...

In modeval: Evaluation of Classification Model Options

Description

suggest_probCut Compare class probability distribution based on probability cut-off. x-axis: probability value. y-axis: observation density.

Usage

suggest_probCut(addTo, outChar, predTag = "pred_test", modelTag = NULL)

Arguments

addTo	Summary list that contains model fits to compare.
outChar	A charactor value of output class name.
predTag	Select prediction results that contains predTag on their name.
modelTag	Select model fits that contains modelTag on their name.

Value

Plot indicating density distribution by probability cut-off.

Examples

## Not run: 
library(mlbench)
data(PimaIndiansDiabetes)
index <- sample(seq_len(nrow(PimaIndiansDiabetes)), 500)
trainingSet <- PimaIndiansDiabetes[index, ]
testSet <- PimaIndiansDiabetes[-index, ]
x <- trainingSet[, -9]
y <- trainingSet[, 9]
x_test <- testSet[, -9]
y_test <- testSet[, 9]
sSummary <- list()
sSummary <- add_model(sSummary, x, y)
sSummary <- add_model(sSummary, x, y, model = c("C5.0Cost", "glmnet"), modelTag = "others")
sSummary <- add_prob(sSummary, x_test, y_test, outChar = "pos")

suggest_probCut(sSummary, "pos")
suggest_probCut(sSummary, "pos", modelTag = "glm|svm")

# vignette("modeval") #check a vignette for further details

## End(Not run)

modeval documentation built on May 29, 2017, 10:54 a.m.