suggest_gain: Plot Gain Chart and Lift Chart

In modeval: Evaluation of Classification Model Options

Description

suggest_gain Gain and Lift charts are widely used in marketing and related contexts. They indicate the effectiveness of predictive models compared to the results obtained without the predictive model.

Usage

suggest_gain(addTo, outChar, predTag = "pred_test", modelTag = NULL,
  cuts = 51, type = 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.
cuts	Integer indicating the number of splits of probability buckets.
type	Plot different type of charts. "Gain" for gain chart. "Lift" for lift chart. "PctAcc" for accumulated event percent. "Pct" for event percent.

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_gain(sSummary, outChar = "pos")
suggest_gain(sSummary, outChar = "pos", modelTag = "glm|svm", type = "Lift")
suggest_gain(sSummary, outChar = "pos", modelTag = "glm|svm", type = "PctAcc")
suggest_gain(sSummary, outChar = "pos", modelTag = "glm|svm", type = "Pct")
suggest_gain(sSummary, outChar = "pos", modelTag = "glm|svm", type = "Gain")
suggest_gain(sSummary, outChar = "pos", modelTag = "glm|svm", type = "Gain") + xlim(0, 0.5)

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

## End(Not run)

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