R Workshop

workshop.r

Getting started
===============
# Make sure you have ggplot2 and plyr installed
install.packages(c("ggplot2", "plyr"), dep=T)

library(ggplot2)
library(plyr)



Basic examples
==============

faithful
head(faithful)
str(faithful)


## Scatter plot

ggplot(data=faithful, mapping=aes(x=eruptions, y=waiting)) + geom_point()

# More concisely:
ggplot(faithful, aes(x=eruptions, y=waiting)) + geom_point()
qplot(eruptions, waiting, data=faithful)


## Histogram

ggplot(faithful, aes(x = eruptions)) + geom_histogram()

ggplot(faithful, aes(x = eruptions)) + geom_histogram(binwidth=.25)


## Points and lines

pressure
str(pressure)

p <- ggplot(pressure, aes(x=temperature, y=pressure))

# Points
p + geom_point()

# Lines
p + geom_line()

# Points with lines
p + geom_line() + geom_point()


# Bar graph (with continous x axis)
p + geom_bar(stat = "identity")






# Two kinds of variables: continuous/numeric vs. discrete/categorical
# ggplot2 treats them very differently!
str(mtcars)
str(pressure)
str(PlantGrowth)



# View the dataset
PlantGrowth
head(PlantGrowth)
str(PlantGrowth)

ggplot(PlantGrowth, aes(x = group, y = weight)) + geom_point()

# Box plot
ggplot(PlantGrowth, aes(x = group, y = weight)) + geom_boxplot()


# View the dataset
PlantGrowth

ggplot(PlantGrowth, aes(x = group, y = weight)) + geom_point()

# Box plot
ggplot(PlantGrowth, aes(x = group, y = weight)) + geom_boxplot()

ggplot(PlantGrowth, aes(x = group, y = weight)) + geom_violin()


Understanding ggplot2
=====================

# All continuous variables
dat <- data.frame(var1 = c(2, 3, 5, 7),
                  var2 = c(2, 4, 8, 5),
                  var3 = c(5, 0, 4, 1))

# With some discrete variables
dat2 <- data.frame(var1 = c("A", "B", "A", "B", "A", "B"),
                   var2 = c("G1", "G0", "G2", "G1", "G0", "G2"),
                   var3 = c(5, 0, 4, 1, 6, 3))

# Mapping data to aesthetics
ggplot(dat, aes(x=var1, y=var2)) + geom_point()
ggplot(dat, aes(x=var1, y=var2, colour=var3)) + geom_point()


# Setting aesthetics
ggplot(dat, aes(x=var1, y=var2)) + geom_point(colour="red")
ggplot(dat, aes(x=var1, y=var2)) + geom_point(colour="red", size=6)


# Different geoms
ggplot(dat, aes(x=var1, y=var2)) + geom_point()
ggplot(dat, aes(x=var1, y=var2)) + geom_line()
ggplot(dat, aes(x=var1, y=var2)) + geom_bar(stat="identity")


# Multiple geoms
ggplot(dat, aes(x=var1, y=var2)) + geom_point() + geom_line()

# Equivalent to 
ggplot(dat) +
    geom_point(aes(x=var1, y=var2)) +
    geom_line(aes(x=var1, y=var2))

ggplot() + 
    geom_point(aes(x=var1, y=var2), data=dat) +
    geom_line(aes(x=var1, y=var2), data=dat)


# Mapping discrete variables
ggplot(dat2, aes(x=var1, y=var3)) + geom_point() 
ggplot(dat2, aes(x=var1, y=var3, colour=var2)) + geom_point()


# Facets
ggplot(dat2, aes(x=var1, y=var3, colour=var2)) + geom_point()
ggplot(dat2, aes(x=var1, y=var3)) + geom_point() + facet_wrap( ~ var2)
ggplot(dat2, aes(x=var1, y=var3)) + geom_point() + facet_grid(. ~ var2)
ggplot(dat2, aes(x=var1, y=var3)) + geom_point() + facet_grid(var2 ~ .)
ggplot(dat2, aes(x=var1, y=var3)) + geom_point() + facet_grid(var1 ~ var2)


# Stats
ggplot(dat2, aes(x=var1, y=var3)) + geom_point()
# Equivalent to
ggplot(dat2, aes(x=var1, y=var3)) + geom_point(stat="identity")
ggplot(dat2, aes(x=var1, y=var3)) + stat_identity(geom="point")

ggplot(dat2, aes(x=var1, y=var3)) + geom_point()




More advanced graphs
=======================

## Scatter plot with regression lines

p <- ggplot(mtcars, aes(x=wt, y=mpg)) + geom_point()
p + geom_smooth()

p + geom_smooth(method=lm, se=FALSE)


## Box plots
ggplot(ToothGrowth, aes(x=supp, y=len)) + geom_point()
ggplot(ToothGrowth, aes(x=supp, y=len)) + geom_boxplot()

## Position adjustments
ggplot(ToothGrowth, aes(x=supp, y=len, fill=factor(dose))) + geom_boxplot()
ggplot(ToothGrowth, aes(x=factor(dose), y=len, fill=supp)) + geom_boxplot()

p <- ggplot(mtcars, aes(x=factor(cyl), fill=factor(am)))
p + geom_bar(position="dodge", colour="black")

p + geom_bar(position="stack", colour="black")


p <- ggplot(mpg, aes(x=displ, y=hwy))
p + geom_point()

p + geom_point(position="jitter")
p + geom_point(position=position_jitter())

p + geom_point(position=position_jitter(width=0.2, height=0))


## Saving

ggplot(mpg, aes(x=displ, y=hwy)) + geom_point()
ggsave("scatter.png")

p <- ggplot(mpg, aes(x=displ, y=hwy)) + geom_point()
ggsave("scatter.png", p, width=4, height=4)

p <- ggplot(mpg, aes(x=displ, y=hwy)) + geom_point()
png("scatter.png")  # Or you can use pdf
print(p)
dev.off()



Wide vs. long data
==================

library(gcookbook)
simpledat
simpledat_long

plum_wide
plum


Group wise operations
=====================

tg_vc<- subset(ToothGrowth, supp == "VC")

mean(tg_vc$len)

# Or
library(plyr)
summarise(tg_vc, mean_len = mean(len))


ddply(ToothGrowth, "supp", summarise, mean_len = mean(len))

# Split on multiple variables
ddply(ToothGrowth, c("supp", "dose"), summarise, mean_len = mean(len))

# Multiple output variables
ddply(ToothGrowth, c("supp", "dose"), summarise, mean=mean(len), sd=sd(len))

# Make a plot
tg <- ddply(ToothGrowth, c("supp", "dose"), 
  summarise, mean=mean(len), sd=sd(len))

ggplot(tg, aes(x=factor(dose), y=mean, fill=supp)) +
    geom_bar(stat="identity", position="dodge", colour="black")



Other resources
===============

# ggplot2 documentation: http://docs.ggplot2.org/current/
# Mailing list: https://groups.google.com/forum/?fromgroups#!forum/ggplot2
# Cookbook for R: http://wiki.stdout.org/rcookbook/
# Stackoverflow: http://stackoverflow.com/questions/tagged/r