# converting into date

setwd('S:/Courses/stat-renaes/Stat404')
nonsales=read.csv('nonsales.csv',header=T)

# look at Birth_Date and Hire_Date

nonsales$Hire_Date
nonsales$Birth_Date

# Notice at the end of the date vectors that it says 
# "Levels: 01/01/1969 01/02/1969 01/02/1978 01/03/1984 01/05/1964 ... 31/12/1978"
# That means that the variable(s) is being treated as a factor (like levels of a
# treatment) rather than numeric or date
# we create new variables by converting the vectors to dates
# we also need date formats
# quotes are needed: 
# "%m/%d/%Y" gives mm/dd/YYYY, etc. (web search for "R as.Date format")
# the dates were entered as dd/mm/YYYY

nonsales$Birthdate=as.Date(nonsales$Birthdate,format="%d/%m/%Y")
nonsales$Hiredate=as.Date(nonsales$Hiredate,format="%d/%m/%Y")

attach(nonsales) 

Birthdate
Hiredate

# no more factors
# now we can calculate differences between the (new) dates and can tell if 
# hiredates are invalid

nonsales$datediff=difftime(Hiredate,Birthdate,units='days')
k=which(nonsales$datediff<0)
nonsales[k,]