##############################
#
# Script: assignment01x.R
#   Graphics for the 
#   answers to the first 
#   homework assignment
#
##############################


# Read in the data

ht <- read.csv("http://courses.kvasaheim.com/pols6123/data/studentHeight.csv")
names(ht)
summary(ht)
attach(ht)


#####
# Problem 1

png("boxplot.png", width=4,height=3,units="in",res=600)
par(mar=c(4,4,1,1)+0.1)
par(cex=0.8,cex.lab=0.8,cex.axis=0.8)
boxplot(height~sex, las=1, xlab="Gender\n", ylab="Height (in)", col=c("#ffaaaa","#aaaaff"))
dev.off()

t.test(height~sex, alternative="greater")


#####
# Problem 2

# Check assumption of Normality using boxplots
boxplot(height~sex, las=1, xlab="Gender", ylab="Height (in)")

# or using histograms
png("histograms.png", width=4,height=3,units="in",res=600)
par(cex=0.8,cex.lab=0.8,cex.axis=0.8)

par(mar=c(0,4,1,1)+0.1)
par(fig=c(0,1,0.55,1))
hist(height[sex=="Male"], main="",xlab="",xlim=c(120,220),xaxt="n", las=1, ylim=c(0,4),col="#aaaaff")
text(130,3,"Males", cex=0.8)

par(mar=c(4,4,0,1)+0.1)
par(fig=c(0,1,0,0.55),new=TRUE)
hist(height[sex=="Female"], main="",xlab="Height (in)\n",xlim=c(120,220), las=1, ylim=c(0,4),col="#ffaaaa")
text(130,3,"Females", cex=0.8)

dev.off()


# or using the Shapiro-Wilk test
shapiro.test(height[sex=="Male"])
shapiro.test(height[sex=="Female"])

# in each of these, Normality seems reasonable 
t.test(height~sex, alternative="greater")