# Intro to R

# What is R?
# Free, open-source, easily-extensible statistical software and programming environment, which runs on all major operating systems.

# Where do I get it?
# http://www.r-project.org
# http://cran.wustl.edu

# Basic syntax
5+3
x <- 5			# a command, giving x the value 5
x = 5			# a command, giving x the value 5
x == 5			# a question, with answers TRUE and FALSE
plot(x=5, y=7)		# specifies that parameter x has value 5
plot(x==5, y=7)		# error

a <- 7
a
a[1]
a[2]
b <- c(2,4,6)
b
b[2]
b[2:3]
b * 4
b + 3
b * b
b[3] <- 100
b
B

C <- matrix(1:4, nrow=2)
C
t(C)
D <- matrix(5:8, nrow=2)
D
C * D
C %*% D
C * 2

exp(5)
exp(b)

e <- c("one", "two", "three")
e[2]

G <- letters
G
G[c(1:10)]
G[-c(1:10)]

unique(c(6,1,2,3,1,2,4,1,2,5))
sort(unique(c(6,1,2,3,1,2,4,1,2,5)))

sample(1:100, 3)


# sample data sets
data(swiss)
swiss
head(swiss)
?swiss
str(swiss)
names(swiss)
edit(swiss)				# Lets you see the data; changes are not saved
fix(swiss)				# Lets you see the data; changes *are* saved to 'swiss'
cleaned.up.swiss <- edit(swiss)		# lets you see the data; changes are saved to 'cleaned.up.swiss'

# Getting help
?exp
?"exp"
?"^"
?"%%"
help.search("exp")
apropos("exp")		# ?foo shows less than apropos("foo") which shows less than help.search("foo")

# RSeek.org
# http://www.rseek.org/  
# Search for "cheat sheet"; gets you: http://cran.r-project.org/doc/contrib/Short-refcard.pdf

# Installing and using packages
install.packages("car")
library(car)
?car

library(foreign)		# Read and write various proprietary data formats, including Stata and SPSS


# Loading data
# Search for "logit"; gets you http://www.ats.ucla.edu/stat/r/dae/logit.htm for third hit.
mydata <- read.csv(url("http://www.ats.ucla.edu/stat/r/dae/logit.csv"))

# Manipulating data

# One and two-dimensional descriptive statistics
summary(swiss)
summary(mydata)
mean(mydata$gre)	# note that we're using the "full name" of the variable
median(mydata$gre)
sd(mydata$gre)

table(mydata$admit)
table(mydata$gre)
hist(mydata$gre)

hist(mydata$gre, prob=TRUE)
lines(density(mydata$gre))

qq.plot(mydata$gpa)		# available because of loading the 'car' package

plot(mydata$gre, mydata$gpa)

# Regression
reg.mydata <- lm(mydata$gpa~mydata$gre)
reg.mydata
summary(reg.mydata)

plot(mydata$gre, mydata$gpa, xlab="GRE", ylab="GPA", main="Observed Relationship Between GRE Score and GPA")
abline(reg.mydata)
abline(reg.mydata, lwd=2, col="green")

# Probit/logit
mydata <- read.csv(url("http://www.ats.ucla.edu/stat/r/dae/logit.csv"))
mylogit<- glm(admit~gre+gpa+topnotch, data=mydata, family=binomial(link="logit"))
summary(mylogit)