library(tidyr)
library(ggplot2)
library(dplyr)
library(plyr)
library(tidyverse) #


#import the data
employee <- read.csv(file="c:\\Users\\Asus\\OneDrive\\Desktop\\employee_attrition.csv",header=TRUE,sep=",")

#different viewing methods
#for first 6 lines
head(employee)
head(employee,10)

#for last 6 lines
tail(employee)
tail(employee,10)

str(employee)
summary(employee)
head(employee)
View(employee)#View the dataset in table form

names(employee)=c("Employee_ID(PK)","Record_Date","Birthday_Date","Orighire_Date","Termination_Date","Age","Length_of_Service","City","Department","Job_Title","Store_Number","Gender","Gender_Full","Term_desc","Type_term","Status_Year","Status","Business_unit")
names(employee)

#Data Cleaning
employee$Gender_Full <- NULL
employee$Store_Number <-NULL
employee$Birthday_Date <-NULL
employee$Type_Term <-NULL
employee$Orighire_Date <- NULL
View(employee)
#how data stores
class(employee)

length(employee)#number of column
ncol(employee)
nrow(employee)#number of row
summary(employee)

#What is the largest in this company
max(employee$Age) #65
#what is the smallest in this company
min(employee$Age) #19
#Who is the longest length of services in this company?
max(employee$Length_of_Service) #26
#Who is the smallest length of services in this company?
min(employee$Length_of_Service) #0

employee[employee$Gender=="Male",] #writing condition with categorical data
male=employee[(employee$Age>60)&(employee$Gender=="M"),]
nrow(male)
View(male)

#How many employee is terminated
Number_Terminated=employee[employee$Status=="TERMINATED",]
#How many employee is worked more than 15 years and terminated
Number_Terminated_1=employee[(employee$`Length of Service`>14)&(employee$Status=="TERMINATED"),]

nrow(Terminated)
View(Terminated)

nrow(Number_Terminated_1)
View(Number_Terminated_1)


Female=employee[(employee$Age>60)&(employee$Gender=="F"),]
nrow(Female)
View(Female)

#Categorize the Length of services
factor(employee$Length_of_Service)#list out the categories
nlevels(factor(employee$Length_of_Service)) #how many level


View(employee)

statusCount<-as.data.frame.matrix(employee %>%
                                    group_by(Status_Year) %>%
                                    select(Status) %>%
                                    table())

statusCount$Total<-statusCount$ACTIVE+statusCount$TERMINATED

statusCount$PercentTerminate<-statusCount$TERMINATED/(statusCount$Total)*100


statusCount

mean(statusCount$PercentTerminate)
#Analysis part
#just terminates

#statusCount<-as.data.frame.matrix(empset %>%
  #                                  group_by(Status) %>%
  #                                 select(STATUS) %>%
#table())
#statusCount$Total<-statusCount$ACTIVE+statusCount$TERMINATED
#statusCount$PercentTerminate<-statusCount$TERMINATED/(statusCount$Total)*100
#statusCount
TerminateData<- employee %>% filter(Status=="TERMINATED")
ggplot(TerminateData)+geom_bar(aes(x=Status_Year, fill=Term_desc))
ggplot(TerminateData)+geom_bar(aes(x=Length_of_Service, fill=Status))
ggplot(TerminateData)+geom_bar(aes(x=as.factor(Department), fill=as.factor(Term_desc)))+
  theme(axis.text.x = element_text(angle=90, hjust=1,vjust=0.5))

AgeData<- employee %>% filter(Age <40 & Status=="TERMINATED")
ggplot(AgeData)+geom_bar(aes(x=Age, fill=Term_desc))
ggplot(data=employee,aes(x=Age,y=Length_of_Service)) + geom_line()
#Scatter plot
#To show older will work how long for the services
ggplot(employee, aes(x=Age,y=Length_of_Service, color=Age)) + geom_point() + xlim(10,60)+ ylim(0,25)
library(caret)
featurePlot(x=MYdataset[,6:7],y=MYdataset$STATUS,plot="density",auto.key = list(columns = 2))
View(employee)