Day 4: In-Class Visualization Exercises.R

# =====================================================================================================================
# VISUALISATION EXERCISES
# =====================================================================================================================

# We'll be using the baseball data from the corrgram package.

# Exercise 1 ----------------------------------------------------------------------------------------------------------

# 1. Install corrgram if necessary.
# 2. Load the library.
# 3. Get acquainted with the baseball data.
# 4. Make a scatter plot showing logSal against Years.
#     - Label axes appropriately. Add a plot title.
#     - Use the col parameter to add a splash of colour.
#     - Change the plotting symbol. Search for "pch" on ?par.
# 5. Generate a pairs plot for the variables `Atbat`, `Hits`, `Homer`, `Runs`, `RBI`, `Walks` and `Years`.
#     - Which of these features are most obviously correlated?

# Exercise 2 ----------------------------------------------------------------------------------------------------------

# 1. Make a box plot of logSal versus Position. Conclusions?
# 2. Make a mosaic plot which shows the distribution of Position versus Team.
#     - See ?spineplot for details of the plot() method.
#     - Create a mosaic plot for Position versus League.

# Exercise 3 ----------------------------------------------------------------------------------------------------------

# 1. Create a histogram for logSal.
#     - Change the bin width to 0.25.
#     - Use a colour to fill the histogram.
#     - Convert from counts to density.

# Exercise 4 ----------------------------------------------------------------------------------------------------------

# Load ggplot2 and use it for the following exercises.

# 1. Make a scatter plot showing logSal against Years.
#     - Colour points by Position. Would this be possible with base graphics? Would it be easy?
#     - Deal with overplotting.
#     - Change the plot theme. I like the classic theme, but you should find something that works for you.
#     - Install ggthemes and explore a wider range of themes.
#     - Add a smoothed curve to the data.
#     - Try a different color palette. See http://colorbrewer2.org/.
#     - Add a 2D density estimate.

# Exercise 5 ----------------------------------------------------------------------------------------------------------

# 1. Make a histogram of logSal.
#     - Sort out any warning messages.
#     - Create overlays for each Position.
#     - Split the plot into facets.
#     - Add apropriate labels for axes. Add a plot title.
# 2. Create a smoothed density estimate with the same data.
#     - Create overlays for each Position.

# Exercise 6 ----------------------------------------------------------------------------------------------------------

# Do something awesome.

# =====================================================================================================================
# VISUALISATION SOLUTIONS
# =====================================================================================================================

# Exercise 1 ----------------------------------------------------------------------------------------------------------

library(corrgram)

plot(logSal ~ Years, data = baseball, xlab = "Years of Play", ylab = "Logarithm of Salary", col = "red", pch = 19)

plot(baseball[, 5:11])

# Exercise 2 ----------------------------------------------------------------------------------------------------------

plot(logSal ~ Position, data = baseball)

plot(Position ~ Team, data = baseball)
mosaicplot(League ~ Position, data = baseball, main = "Mosaic: Position verus League")

# Exercise 3 ----------------------------------------------------------------------------------------------------------

hist(baseball$logSal)
hist(baseball$logSal, breaks = seq(1.75, 3.5, 0.25))
hist(baseball$logSal, breaks = seq(1.75, 3.5, 0.25), col = "lightgreen")
hist(baseball$logSal, breaks = seq(1.75, 3.5, 0.25), col = "lightgreen", probability = TRUE)

# Exercise 4 ----------------------------------------------------------------------------------------------------------

library(ggplot2)

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point()

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position))

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position), alpha = 0.5, position = "jitter")

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position), alpha = 0.5, position = "jitter") +
  theme_classic()

library(ggthemes)

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position), alpha = 0.5, position = "jitter") +
  theme_wsj()

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position), alpha = 0.5, position = "jitter") +
  geom_smooth() + scale_colour_brewer(palette = "Set1") +
  theme_classic()

ggplot(baseball, aes(x = Years, y = logSal)) +
  geom_point(aes(col = Position), alpha = 0.5, position = "jitter") +
  geom_density_2d(colour = "darkgrey") +
  geom_smooth() + scale_colour_brewer(palette = "Set1") +
  theme_classic()

# Exercise 5 ----------------------------------------------------------------------------------------------------------

ggplot(baseball, aes(x = logSal)) +
  geom_histogram(binwidth = 0.125) +
  theme_classic()

# Stacked version
#
ggplot(baseball, aes(x = logSal)) +
  geom_histogram(aes(fill = Position), binwidth = 0.125) +
  theme_classic()

ggplot(baseball, aes(x = logSal)) +
  geom_histogram(aes(fill = Position), binwidth = 0.125, alpha = 0.55, position = "identity") +
  theme_classic()

ggplot(baseball, aes(x = logSal)) +
  geom_histogram(aes(fill = Position), binwidth = 0.125, alpha = 0.55) +
  facet_grid(Position ~ .) +
  labs(x = "Logrithm of Salary", y = "Count") +
  ggtitle("Salary Distribution by Position") +
  theme_classic()

ggplot(baseball, aes(x = logSal)) +
  geom_density(aes(fill = Position), color = "black", alpha = 0.5) +
  scale_fill_brewer(palette = "Set1") +
  theme_classic()

# Exercise 6 ----------------------------------------------------------------------------------------------------------