DensityPlot: Density Plot (Plot)

In mariytu/RegressionLibs: What the package does (short line)

Description

Generate a density plot for a specific column of the data.

Usage

DensityPlot(data, col)

Arguments

data	an object of class data frame with the data.
col	an integer that specify the column that you want for make the plot.

See Also

http://www.rdatamining.com/examples/outlier-detection

Examples

#Example 1
#install.packages("Rlof")
library(Rlof) #for outlier detection

iris.x <- iris[,1:4] # These are the independent variables
Species <- iris[,5] # This is the dependent variable

DensityPlot(iris.x,1)


#Example 2
#install.packages("Rlof")
library(Rlof) #Outlier detection library

iris.x <- iris[,1:4] # These are the independent variables
Species <- iris[,5] # This is the dependent variable

outlier.scores <- lof(iris.x, k = 5) #applying outlier detection
outlier.scores <- data.frame(outlier.scores)
DensityPlot(outlier.scores, 1) #Generating a plot of outliers scores


#Example 3
#install.packages("Rlof")
library(Rlof) #Outlier detection library

iris.x <- iris[,1:4] # These are the independent variables
Species <- iris[,5] # This is the dependent variable

outlier.scores <- lof(iris.x, k = c(5:10)) #applying outlier detection
mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores


#Example 4
#install.packages("Rlof")
library(Rlof) #Outlier detection library
library(plyr)
# Getting a data set without missing values
cars <- read.csv("https://dl.dropboxusercontent.com/u/12599702/autosclean.csv", sep = ";", dec = ",")

cars.x <- cars[,1:16] # These are the independent variables
cars.y <- cars[,17] # This is the dependent variable

outlier.scores <- lof(cars.x, k = c(5:10)) #applying outlier detection
mean <- rowMeans(outlier.scores) #Calculating the mean of every execution
outlier.scores <- data.frame(outlier.scores, mean) #adding mean to data frame
DensityPlot(outlier.scores, ncol(outlier.scores)) #Generating a plot of outliers scores

aux <- outlier.scores[,7]>1.7 #1.7 is the threshold selected
count(aux)[2,2] #Number of outliers found
outliers <- order(outlier.scores[,7], decreasing=T)[1:count(aux)[2,2]] #Getting the values that are on the threshold
Score <- outlier.scores[outliers,7] #Getting outliers scores
outliers <- data.frame(outliers,Score)
names(outliers) <- c("Position","Score")
View(outliers)

auxOutliers <- outlier.scores[-outliers[1:3,1],] #Eliminating the 3 most remote instances!
DensityPlot(auxOutliers, ncol(outlier.scores)) #Generating a plot of outliers scores

mariytu/RegressionLibs documentation built on May 21, 2019, 11:47 a.m.