library(mlbench)
#data(package="mlbench")
data(Ozone)
head(Ozone)
names(Ozone) <- c("month","daym","dayw","ozone","pressure","windspeed",
	"humidity","tempS","tempEM","inversion","grad","inverstemp","visibility")

head(Ozone)
#ozone <- Ozone[,4]
ozone <- Ozone$ozone

hist(ozone,main="",density=10,xlab="Daily maximum one-hour-average ozone reading",cex.lab=1.2,breaks=15)
hist(log(ozone),main="",density=10,xlab="Log-daily maximum one-hour-average ozone reading",cex.lab=1.2,breaks=15)

ozone <- na.omit(ozone)
density.dat <- density(ozone)
plot(density.dat)
plot(density(ozone))

plot(Ozone$tempS,Ozone$ozone,xlab="TempS",ylab="Ozone",pch=16)

Ozone.nomissing <- na.omit(Ozone)
is.data.frame(Ozone.nomissing)
D <- as.data.frame(Ozone.nomissing)
head(D)

lines(lowess(Ozone.nomissing$ozone~Ozone.nomissing$tempS))

lm.obj <- lm(log(ozone)~pressure+tempS+grad+inverstemp,data=Ozone.nomissing)
summary(lm.obj)
plot(lm.obj)


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

library(DAAG)
data(ais)

ais$sport
here <- ais$sport %in%  c("Row","Swim")

# load lattice package from drop-down menu before proceeding

library(lattice)


xyplot(ssf~ht | sport,groups=sex,aspect=1,auto.key=list(columns=2),subset=here,data=ais,
	xlab="Height",ylab = "Skinfold thickness",pch=16)

xyplot(ais$pcBfat~ais$ssf|ais$sport,groups=ais$sex,col=c("black","red"),pch=16,
	xlab="Skinfold thickness",ylab="Percent body fat",
	key = list(corner=c(.8,.9),cex=.8,cex.title=1.05,points=T,
                 pch=16,col=c("black","red"),border=F,
                  text = list(lab = c("Females","Males"),
                  columns = 1)))

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

data <- read.table("C:/Documents and Settings/steeleb/My Documents/Math 542/BrainWeight.txt",sep=",",header=TRUE)
plot(data)

brain <- log(data$brain)
body <- log(data$body)
species <- data$species

plot(brain~body,xlab="Log(body)",ylab="Log(brain)")
text(brain~body,labels=species,pos=4)

plot(brain~body,xlab="Log(body)",ylab="Log(brain)")
primates.index <- 6:24
points(body[primates.index],brain[primates.index],col=2,pch=16)
identify(brain~body,labels=species)

lines(lowess(body,brain,f=.2),col=2)
lines(lowess(body,brain,f=.1),col=3)

lines(lowess(body,brain,f=.5),col=4)

lm.obj <- lm(brain~body)
abline( lm.obj$coeff)

summary(lm.obj)
options(digits=3)
summary(lm.obj)



# vectors, matrices and factors

x <- c(1,2,3)
x

2*x
mean(x)
sd(x)
y <- c(T,T,F,T,T)
name <- c("Ape","Monkey")
name

x2 <- 1:5
x2
c(x,x2)
x2[1:3]
rbind(x,x2)
M <- rbind(x,x2[1:3])
M
is.matrix(M)

I <- diag(3)
I

I[1,]%*%x
I[2,]%*%x
I[3,]%*%x

j <- 1
while(j < 4) {
	print(I[j,]%*%x)
	j <- j + 1
}
I%*%x


########################## missing values ######

Ozone$grad
Ozone$grad==NA  # one missing value makes all values missing
mean(Ozone$grad)
grad <- na.omit(Ozone$grad)
mean(grad)
mean(Ozone$grad,na.rm=TRUE)

Ozone <- na.omit(Ozone)

n <- dim(Ozone)[1]
X <- matrix(unlist(Ozone[,c(1,2,3,5,6,7,8,10)]),n,8)
X <- cbind(rep(1,n),X)
y <- Ozone[,4]
b <- solve(t(X)%*%X,t(X)%*%y)
mean((y - X%*%b)^2)

n <- 100
s <- sample(1:10,n,replace=TRUE)
s