# Reading data
x <- read.csv(file.choose())  # Open tips data
x <- read.table(file.choose(), sep=",", header=T)
head(x)
x <- read.table(file.choose(), sep=",", header=T, row.names=1)
head(x)
x <- scan(file.choose(), what=character(), sep=",")
dim(x)
length(x)
head(x,50)

data()
data(package="datasets")
data(AirPassengers)
head(AirPassengers)
?AirPassengers

# Data structures
head(sla)
is.data.frame(sla)
sla[[1]]
sla$me
sla[,1]
sla[[1]][1]
sla$me[1]
sla[1,1]
sla[[2]]
str(sla)

# Your turn 1
sla[[3]][10]
sla[10,3]
sla$month[10]
str(sla)

# Data structures 2
A<-matrix(c(1,2,-2,1),ncol=2,byrow=T)
A
x<-matrix(rnorm(10),ncol=2)
x
x%*%A
x%*%t(A)
var(x)
prcomp(var(x))
apply(x,2,mean)

# Types of variables
str(sla)
is.numeric(sla$year)
as.character(sla$year)
sla$me
as.numeric(sla$me)

is.factor(sla$me)
levels(sla$me)
x<-factor(sla$me, levels=c("94s10", "007angie", "other"))
levels(x)
x[1:50]
x[is.na(x)]<-"other"
x[1:50]

# Missing values
# Read in the shangri-la person level data
head(slap)
summary(slap)
mean(slap$age)
mean(slap$age, na.rm=T)
age2 <- slap$age
xm <- mean(slap$age, na.rm=T)
age2[is.na(age2)] <- xm
mean(age2)
sd(age2)
sd(slap$age, na.rm=T)

# Your turn 2
is.factor(sla$me)
levels(sla$me)
x<-factor(sla$me, levels=c("splitbamboo", "rhonda", "other"))
levels(x)
x[1:50]
x[is.na(x)]<-"other"
x[1:50]

wgt2 <- slap$target.weight
xm <- median(slap$target.weight, na.rm=T)
wgt2[is.na(wgt2)] <- xm
mean(wgt2)
sd(wgt2)
sd(slap$target.weight, na.rm=T)

# Subsetting
slap.sub <- slap[slap$target.weight < 145.0,]
dim(slap.sub)
dim(slap)
summary(slap.sub)

sla.sub <- sla[sla$me=="millifoo",]
dim(sla.sub)
sla.sub

slap.ord <- slap[order(slap$target.weight),]
head(slap.ord)