#####
# lecture 1
####

setwd("I:\\Classes\\Regression 2009\\data")
data <- read.csv("arsenic.csv")
par(bg="white")
boxplot(arsnails, arswater, names=c(" Nails","Water"), ylab="Arsenic, ppm")


par(mfrow=c(1,2))
boxplot(data$arsnails,  names=c("Nails"), ylab="Arsenic in ppm")
boxplot(data$arswater,  names=c("Water"), ylab="Arsenic in ppm")

par(mfrow=c(2,1))
par(mar=c(5,4,1,1))
hist(data$arsnails, xlab="Arsenic in Nails, ppm", main="", breaks=8)
hist(data$arswater, xlab="Arsenic in Water, ppm", main="", breaks=3)

plot(density(data$arsnails),  xlab="Arsenic in Nails, ppm", main="", 
	ylab="Density")
plot(density(data$arswater),  xlab="Arsenic in Water, ppm", main="", 
	ylab="Density")

par(mfrow=c(1,1))
plot(jitter(data$sex), data$arsnails, xlim=c(0.75,2.25), xaxt="n",
	ylab="Arsenic, ppm", xlab="", pch=16, cex=1.2)
axis(1, at=c(1,2),c("Male","Female"))


y <- log(10000*arsnails)
x <- log(10000*arswater+1)

par(mfrow=c(1,1))
par(bg="white")
plot(data$arswater, data$arsnails, ylab="Level of Arsenic in Nails (ppm)",
		xlab="Level of Arsenic in Well Water (ppm)", pch=1, cex=1.3)
reg <- lm(y ~ x)
abline(reg, lwd=2)
detach(2)

#################################################

data <- read.csv("fatherson.csv")
data <- data[1:200,]
attach(data)

par(bg="white")
plot(father, son, xlab="Father's Height, Inches", ylab="Son's Height, Inches",
	xaxt="n",yaxt="n",ylim=c(58,78), xlim=c(58,78))
axis(1, at=seq(58,78,2))
axis(2, at=seq(58,78,2))

reg <- lm(son~father)
summary(reg)

abline(reg, lwd=2 )
abline(0,1, col=2, lwd=2)
legend(70,62, c("x=y","regression line"), col=c(2,1), lwd=c(2,2))

mean(father)
mean(son)