suggest_probPop: Compare class probability distributions based on the...

In modeval: Evaluation of Classification Model Options

Description

suggest_probPop Compare class probability distribution based on the population of observations, sorted by probability values. x-axis: percentile of class probability. y-axis: observation density.

Usage

suggest_probPop(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 observed population.

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_probPop(sSummary, "pos")
suggest_probPop(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.