library(foreign)
setwd("Z:/Econometrie")

MyData <- read.spss('FESYOUTHSTUDYROMANIA.sav',to.data.frame=TRUE)
View(MyData)

backupdata=MyData

grade<- MyData$e4

hist(grade) # makes a histogram over the Height.Inches. variable
mean(grade)
median(grade)
sd(grade)
variance(grade)

#Crosstabulation for two variables
MyData.ctab<- table(MyData$dir,MyData$tenyr)
MyData.ctab
margin.table(MyData.ctab) # see total observations
round(prop.table(MyData.ctab),4) # computes proportion with 4 decimals

crosstab<-as.data.frame (round(prop.table(MyData.ctab),2)) # show proportion with 2 decimals into a table
View(crosstab) # see the table

# creating a data frame with the name of the variables
attr(MyData, 'variable.labels')
MyData.labels <- as.data.frame(attr(MyData, 'variable.labels'))

#recode variables
MyData$CopyOfc13_5 <- NA


#create new data set with the research variables
myvar<-subset(MyData, select= c("e4", "a7_5", "a7_1", "a7_2", "tenyr"))

#remove NA value
complete.cases(myvar) # shows true (if there is a value) or false (if there is not any value)
sum(which (complete.cases(myvar))) #all the complete values
sum(which (!complete.cases(myvar))) # all the incomplete values (! stands for not)
research<- na.omit(myvar)
sum(which (!complete.cases(research))) # if it is 0 then there are no NA values
View(research)

#check for other missing values in variable tenyr
table(research$tenyr)

# remove DK/NA values
research[research$tenyr == 'DK/NA', 'tenyr']<- NA
table(research$tenyr)
sum(which (!complete.cases(research))) # check for NA values in tenyr
research<- na.omit(research)  # remove NA values in tenyr
sum(which (!complete.cases(research))) #chek if NA values in tenyr are 0
table(research$tenyr)
View(research)

#transform factor data as numerical data
research$e4<-as.numeric(research$e4) # average grade last year
research$a7_5<-as.numeric(research$a7_5) # money spend on books
research$a7_1<-as.numeric(research$a7_1) # money spend on films
research$a7_2<-as.numeric(research$a7_2) #money spend on bars


# simple linear regression model
lm(e4 ~ a7_5, data=research)

#multiple regression model
LRM<-lm(e4 ~ a7_5+a7_2+a7_1, data=research)
LRM # to see the results of the equation
View(LRM)
summary(LRM) # to see the descriptive statistics and the equation model

install.packages("lmtest") # after installation you need to go to packages and check the lmtest package to be functional or use library
library (lmtest)
dwtest(LRM)