Outliers

Data

ID<-c(16762,16439,16211,16790,16443,16998,
      16543,16779,16945,16111,16224,16980,
      16779,16000,16111,16224,16400,16327)
Name<-c("Ahmed","Osama","Ibraheem","Fahd",
        "Majeda","Hdeel","Mohammed","Remas",
        "Rteel","Abdalrhman","Mhdi","Tala",
        "Remas","Nadiah","Abdalrhman","Mhdi",
        "Lila","Fatima")
Age<-c(10,2,2,27,9,32,9,10,29,28,9,
       9,30,29,28,6,33 ,100)
Sex<-c("M","M","M","M","F","F","M","F",
       "F","M","M","F","F","F","M","M",
       "F","F")

data<-data.frame(ID,Name,Age,Sex)
data

##       ID       Name Age Sex
## 1  16762      Ahmed  10   M
## 2  16439      Osama   2   M
## 3  16211   Ibraheem   2   M
## 4  16790       Fahd  27   M
## 5  16443     Majeda   9   F
## 6  16998      Hdeel  32   F
## 7  16543   Mohammed   9   M
## 8  16779      Remas  10   F
## 9  16945      Rteel  29   F
## 10 16111 Abdalrhman  28   M
## 11 16224       Mhdi   9   M
## 12 16980       Tala   9   F
## 13 16779      Remas  30   F
## 14 16000     Nadiah  29   F
## 15 16111 Abdalrhman  28   M
## 16 16224       Mhdi   6   M
## 17 16400       Lila  33   F
## 18 16327     Fatima 100   F

1)Use box plot

box_plot <-boxplot(data$Age)$out
mtext(paste("Outliers: ", paste(box_plot, collapse = ", ")))

#Identify rows containing outliers
out_ind <- which(data$Age %in% c(box_plot))
out_ind

## [1] 18

data[out_ind, ]

##       ID   Name Age Sex
## 18 16327 Fatima 100   F

#Remove outliers 
outliers <- boxplot(data$Age, plot=FALSE)$out
outliers

## [1] 100

data[-which(data$Age %in% outliers),]

##       ID       Name Age Sex
## 1  16762      Ahmed  10   M
## 2  16439      Osama   2   M
## 3  16211   Ibraheem   2   M
## 4  16790       Fahd  27   M
## 5  16443     Majeda   9   F
## 6  16998      Hdeel  32   F
## 7  16543   Mohammed   9   M
## 8  16779      Remas  10   F
## 9  16945      Rteel  29   F
## 10 16111 Abdalrhman  28   M
## 11 16224       Mhdi   9   M
## 12 16980       Tala   9   F
## 13 16779      Remas  30   F
## 14 16000     Nadiah  29   F
## 15 16111 Abdalrhman  28   M
## 16 16224       Mhdi   6   M
## 17 16400       Lila  33   F

Data

ID<-c(16762,16439,16211,16790,16443,16998,
      16543,16779,16945,16111,16224,16980,
      16779,16000,16111,16224,16400,16327)
Name<-c("Ahmed","Osama","Ibraheem","Fahd",
        "Majeda","Hdeel","Mohammed","Remas",
        "Rteel","Abdalrhman","Mhdi","Tala",
        "Remas","Nadiah","Abdalrhman","Mhdi",
        "Lila","Fatima")
Age<-c(10,2,2,27,9,32,9,10,29,28,9,
       9,30,29,28,6,33 ,100)
Sex<-c("M","M","M","M","F","F","M","F",
       "F","M","M","F","F","F","M","M",
       "F","F")

data<-data.frame(ID,Name,Age,Sex)
data

##       ID       Name Age Sex
## 1  16762      Ahmed  10   M
## 2  16439      Osama   2   M
## 3  16211   Ibraheem   2   M
## 4  16790       Fahd  27   M
## 5  16443     Majeda   9   F
## 6  16998      Hdeel  32   F
## 7  16543   Mohammed   9   M
## 8  16779      Remas  10   F
## 9  16945      Rteel  29   F
## 10 16111 Abdalrhman  28   M
## 11 16224       Mhdi   9   M
## 12 16980       Tala   9   F
## 13 16779      Remas  30   F
## 14 16000     Nadiah  29   F
## 15 16111 Abdalrhman  28   M
## 16 16224       Mhdi   6   M
## 17 16400       Lila  33   F
## 18 16327     Fatima 100   F

2)Use inter quartile range

Upper Range = Q3+1.5IQR
Lower Range = Q1-1.5
Outliers = Observations > Q3 + 1.5IQR or < Q1 – 1.5*IQR

Q1<-quantile(data$Age,.25);Q1

## 25% 
##   9

Q3<-quantile(data$Age,.75);Q3

## 75% 
##  29

IQR<-IQR(data$Age);IQR

## [1] 20

#Upper Range Q3+1.5*IQR
Up<-29+(1.5*IQR)
#Lower Range Q1-1.5*IQR
Low<-9-(1.5*IQR)

#Remove outliers 
subset(data,data$Age>(9-(1.5*IQR))&data$Age<(29+(1.5*IQR)))

##       ID       Name Age Sex
## 1  16762      Ahmed  10   M
## 2  16439      Osama   2   M
## 3  16211   Ibraheem   2   M
## 4  16790       Fahd  27   M
## 5  16443     Majeda   9   F
## 6  16998      Hdeel  32   F
## 7  16543   Mohammed   9   M
## 8  16779      Remas  10   F
## 9  16945      Rteel  29   F
## 10 16111 Abdalrhman  28   M
## 11 16224       Mhdi   9   M
## 12 16980       Tala   9   F
## 13 16779      Remas  30   F
## 14 16000     Nadiah  29   F
## 15 16111 Abdalrhman  28   M
## 16 16224       Mhdi   6   M
## 17 16400       Lila  33   F