############################################################
## R code to reproduce statistical analysis in the textbook:
## Agresti, Franklin, Klingenberg
## Statistics: The Art & Science of Learning from Data
## 4th Edition, Pearson 2017
## Web: ArtofStat.com
## Copyright: Bernhard Klingenberg
###########################################################

###################
###  Chapter 2  ###
###  Example 7  ###
###################

###############
## Histogram ##
###############

# Read in Sodium values:
Sodium <- c(0, 340, 70, 140, 200, 180, 210, 150, 100, 130, 140, 180, 190, 160,290, 50, 220, 180, 200, 210)

# Create Basic Histogram:
hist(Sodium, xlab="Sodium (mg)", ylab="Frequency", main="Distribution of Sodium Values in Cereals", col="tan")

# Changing the bins by providing the boundaries.
# (Note: right=FALSE puts an observation such as 120 in the interval from 120-160 and not 80-120)
hist(Sodium, breaks=seq(0,360,40), right=FALSE, xlab="Sodium (mg)", ylab="Frequency", main="Distribution of Sodium Values in Cereals", col="tan")

# Another way to request a certain number of bins:
hist(Sodium, breaks=10, xlab="Sodium (mg)", ylab="Frequency", main="Distribution of Sodium Values in Cereals", col="tan")

# Too few breaks:
hist(Sodium, breaks=2, xlab="Sodium (mg)", ylab="Frequency", main="Distribution of Sodium Values in Cereals", col="tan")

# Too many breaks:
hist(Sodium, breaks=30, xlab="Sodium (mg)", ylab="Frequency", main="Distribution of Sodium Values in Cereals", col="tan")

# For more fine tuning, it is better to use the ggplot2 library.
# If you haven't installed it already, first type: install.packages(ggplot2)
library(ggplot2)
# Adjusting x-axis labels:
ggplot(data.frame(Sodium), aes(x=Sodium)) + 
  geom_histogram(breaks=seq(0,360,40), color="black", fill="tan") +
  labs(x="Sodium (mg)", y="Frequency",  
       title="Distribution of Sodium Values in Cereals") +
  scale_x_continuous(breaks=seq(0,360,40)) +
  scale_y_continuous(limit=c(0,7), breaks=1:7, expand=c(0,0)) +
  theme_bw() +
  theme(panel.grid.minor=element_blank(), text=element_text(size=14))
  

# Plotting percentages rather than counts on the y-axis:
ggplot(data.frame(Sodium), aes(x=Sodium, y=100*(..count../sum(..count..)))) + 
  geom_histogram(breaks=seq(0,360,40), color="black", fill="tan") +
  labs(x="Sodium (mg)", y="Percent (%)",  
       title="Distribution of Sodium Values in Cereals") +
  scale_x_continuous(breaks=seq(0,360,40)) +
  scale_y_continuous(limit=c(0,32), breaks=seq(0,30,5), expand=c(0,0)) +
  theme_bw() +
  theme(panel.grid.minor=element_blank())

# R actually defines intervals open to the left and closed to the right 
# To get the histograms perfectly match the ones in the textbook, use closed="left":
ggplot(data.frame(Sodium), aes(x=Sodium, y=100*(..count../sum(..count..)))) + 
  geom_histogram(breaks=seq(0,360,40), closed="left", color="black", fill="tan") +
  labs(x="Sodium (mg)", y="Percent (%)",  
       title="Distribution of Sodium Values in Cereals") +
  scale_x_continuous(breaks=seq(0,360,40)) +
  scale_y_continuous(limit=c(0,27), breaks=seq(0,25,5), expand=c(0,0)) +
  theme_classic() +
  theme(panel.grid.minor=element_blank())