Learning Objectives

In the Asynchronous Lecture

Understanding the components of a good quantitative visualization
An introduction to the grammar of graphics (ggplot2)
Delving into the different geometric projections available to you when visualizing:
- continuous on continuous data;
- continuous on categorical data; and
- categorical on categorical data.
Tweaking aesthetics with themes and scales.
Combining many plots with faceting and patchwork

In the Synchronous Lecture

Further develop our understanding of the ggplot2 package with an applied walkthrough
Discuss how to export ggplot2 graphics.
Discover some third-party packages that make specific visualization tasks easier.

If you have any questions while watching the pre-recorded material, be sure to write them down and to bring them up during the synchronous portion of the lecture.

Asynchronous Materials

The following tabs contain pre-recorded lecture materials for class this week. Please review these materials prior to the synchronous lecture.

Total time: Approx. 1 hour and 12 minutes

_

Visualizing Data

Grammar of Graphics

Code from the video

require(tidyverse)

# Data ------------------------------------------------------------------

# Simulated Data
set.seed(1234)
dat <- 
  tibble::tibble(var1 = rnorm(100),
                 var2 = var1 + rnorm(100))

summary(dat)


# Grammar of Graphics --------------------------------------------------
ggplot(data = dat, aes(x = var1, y = var2, color = var1)) +
  geom_point()

Continuous Data

Code from the video

require(tidyverse)

# Data --------------------------------------------------------------------

# Simulated Data
set.seed(1234)
dat <- 
  tibble::tibble(var1 = rnorm(1000),
                 var2 = var1 + rnorm(1000))

summary(dat)

dat


# Continuous Univariate -----------------------------------------------------

# Histogram
ggplot(data = dat, aes(x = var1)) +
  geom_histogram()


# Density Plot
ggplot(data = dat, aes(x = var1)) +
  geom_density(fill="blue",color="white",alpha=.5)

# Continuous Bivariate ----------------------------------------------------

# Scatter Plot
ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_point()


# Line Plot
ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_line()


# Density Plot (in 2 Dimensions)
ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_density_2d()

# Same thing filled
ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_density_2d_filled()

# Hex Plot
ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_hex()


# Layering Geoms ----------------------------------------------------

ggplot(data = dat, aes(x = var1, y = var2)) +
  geom_point() +
  geom_hex(alpha=.7)

Categorical Data

Code from the video

require(tidyverse)
library(palmerpenguins) # https://github.com/allisonhorst/palmerpenguins

    # Includes measurements for penguin species, island in Palmer Archipelago,
    # size (flipper length, body mass, bill dimensions), and sex. This is a
    # subset of penguins_raw.


# Data --------------------------------------------------------------------

dat <- penguins

summary(dat)



# Univariate Categorical --------------------------------------------------


# Bar Plot
ggplot(dat,aes(x=species)) +
  geom_bar()


# Ordering Bar Plot by Frequency
ggplot(dat,aes(x=fct_infreq(species))) +
  geom_bar()


# Adding in more categorical data 
ggplot(dat,aes(x=fct_infreq(species), fill = sex)) +
  geom_bar()


# Stacking vs. Dodge
ggplot(dat,aes(x=fct_infreq(species), fill = sex)) +
  geom_bar(position="dodge")



# Bivariate: category on continuous -----------------------------------------


# Box plot
ggplot(dat,aes(x=body_mass_g,y = species)) +
  geom_boxplot()


# Violin plot
ggplot(dat,aes(x=body_mass_g,y = species)) +
  geom_violin()


# Jitter plot
ggplot(dat,aes(x=body_mass_g,y = species)) +
  geom_jitter(height = .05,alpha=.5)


# Layer the representations
ggplot(dat,aes(x=body_mass_g,y = species)) +
  geom_jitter(height = .15) +
  geom_violin(alpha=.5)


# Trivariate: category on category on continous -----------------------------------------

dat %>% 
  drop_na() %>% 
  ggplot(aes(x=sex,y = species,fill=body_mass_g)) +
  geom_tile()


# Trivariate: category on category on continous -----------------------------------------

dat %>% 
  ggplot(aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
  geom_point()

Many Plots

Code from the video

require(tidyverse)
require(patchwork) # for combining plots
require(palmerpenguins) # https://github.com/allisonhorst/palmerpenguins

    # Includes measurements for penguin species, island in Palmer Archipelago,
    # size (flipper length, body mass, bill dimensions), and sex. This is a
    # subset of penguins_raw.


# Data --------------------------------------------------------------------

dat <- penguins

summary(dat)


# Many Plots with Faceting -----------------------------------------


  ggplot(dat,
         aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
    geom_point() 


  # Break plots up by category with facets
  ggplot(dat,
         aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
    geom_point() +
    facet_wrap(~species)
  
  
  # Adjust scales on the facets
  ggplot(dat,
         aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
    geom_point() +
    facet_wrap(~species,scales = "free")
  
  
  # Facet along more than one category
  ggplot(drop_na(dat),
         aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
    geom_point() +
    facet_wrap(~species + sex)
  
  
  # Specify the columns and rows
  ggplot(drop_na(dat),
         aes(x=bill_length_mm,
             y = flipper_length_mm,
             color=species)) +
    geom_point() +
    facet_wrap(~species + sex,scales = "free",ncol = 2)
  


# Combine Plots with Patchwork --------------------------------------------

  plt1 <- 
    ggplot(dat,aes(x=bill_length_mm,
                   y=flipper_length_mm)) +
    geom_point()
  plt1
  
  plt2 <- 
    ggplot(dat,aes(x=fct_infreq(species), fill = sex)) +
    geom_bar(position="dodge")
  plt2
  
  
  # Combine them with patchwork
  plt1 + plt2

  # Arrange them
  plt1 + plt2 + plot_layout(ncol = 1)
  
  
  # resize them
  plt1 + plt2 + plot_layout(ncol = 1,heights = c(.25,.75))
  
  
  # sky is the limit
  plt1 + plt2 + plt2 + plt1

Customizing Plots

Code from the video

require(tidyverse)
require(palmerpenguins) # https://github.com/allisonhorst/palmerpenguins

    # Includes measurements for penguin species, island in Palmer Archipelago,
    # size (flipper length, body mass, bill dimensions), and sex. This is a
    # subset of penguins_raw.


# Data --------------------------------------------------------------------

dat <- penguins

summary(dat)


# Customizing Plot Aesthetics -----------------------------------------

ggplot(dat,
       aes(x=bill_length_mm,
           y = flipper_length_mm,
           color=species)) +
  geom_point(size=3,alpha=.75) +
  scale_color_manual(values = c("darkred","steelblue","grey30")) +
  theme_classic() +
  labs(x = "Bill Length (mm)",y="Flipper Length (mm)",color="") +
  theme(legend.position = "top")

Practice

These exercises are designed to help you reinforce your grasp of the concepts covered in the asynchronous lecture material.

For the following questions, we’ll use the diamonds dataset, which comes installed when we import the ggplot2 package. The dataset contains the prices and other attributes of 54,000 diamonds. The dataset is useful for visualization purposes.

require(tidyverse)
dat <- diamonds 
head(dat)

## # A tibble: 6 x 10
##   carat cut       color clarity depth table price     x     y     z
##   <dbl> <ord>     <ord> <ord>   <dbl> <dbl> <int> <dbl> <dbl> <dbl>
## 1 0.23  Ideal     E     SI2      61.5    55   326  3.95  3.98  2.43
## 2 0.21  Premium   E     SI1      59.8    61   326  3.89  3.84  2.31
## 3 0.23  Good      E     VS1      56.9    65   327  4.05  4.07  2.31
## 4 0.290 Premium   I     VS2      62.4    58   334  4.2   4.23  2.63
## 5 0.31  Good      J     SI2      63.3    58   335  4.34  4.35  2.75
## 6 0.24  Very Good J     VVS2     62.8    57   336  3.94  3.96  2.48

_

Question 1

Using the diamonds data, plot the distribution of the different cut categories. Customize the plot as follows:

Change the fill of the bars be "steelblue";
Change the theme of plot to be minimal;
Add a title using labs() that reads “Diamond Cuts”

_

Answer

ggplot(dat,aes(x=cut)) +
  geom_bar(fill='steelblue') +
  labs(title = "Diamond Cuts") +
  theme_minimal()

Question 2

Do the following:

Create a new variable in the diamonds data called expensive that takes on the value of "yes" if the price is above the average price of a diamond, "no" otherwise.
Using the same code you used to generate the plot in Q1, use this variable to break up each category into expensive and non-expensive stacks using the "dodge" position (see the video on categorical data).
When expensive == "yes" the bars should be filled orange, and grey30 otherwise.

_

Answer

dat %>% 
  mutate(expensive = ifelse(price > mean(price),"yes","no")) %>% 
  ggplot(aes(x=cut,fill=expensive)) +
  geom_bar(position="dodge") +
  scale_fill_manual(values = c("grey30","orange")) +
  labs(title = "Diamond Cuts") +
  theme_minimal()

Question 3

Using the diamonds data, plot price on carat using points, and do the following:

Make the points transparent using alpha = .25;
Color each point by cut.
Facet the plots by cut and arrange the facet plots so that you only have 1 row;
Change the labels so that the legend reads “Diamond Cut”, the x-axis reads “Price ($)” and the y-axis reads “Carat”.
Change the theme to classic;
Change the position of the legend so that it’s on the "bottom" of the plot.

_

Answer

dat %>% 
  ggplot(aes(price, carat, color= cut)) +
  geom_point(alpha=.25) +
  facet_wrap(~cut,nrow=1) +
  labs(color="Diamond Cut",x = "Price ($)",y="Carat") +
  theme_classic() +
  theme(legend.position = "bottom")

The following materials were generated for students enrolled in PPOL670. Please do not distribute without permission.

ed769@georgetown.edu | www.ericdunford.com

Back to Course Website

Data Visualization

PPOL 670 | Introduction to Data Science

Lecture Materials for Week 5

Professor Eric Dunford (ed769@georgetown.edu)
McCourt School of Public Policy, Georgetown University

Learning Objectives

Asynchronous Materials

_

Visualizing Data

Relevant Slides

Grammar of Graphics

Relevant Slides

Code from the video

Continuous Data

Code from the video

Categorical Data

Code from the video

Many Plots

Code from the video

Customizing Plots

Code from the video

Practice

_

Question 1

_

Answer

Question 2

_

Answer

Question 3

_

Answer

Back to Course Website Data Visualization

PPOL 670 | Introduction to Data Science Lecture Materials for Week 5

Professor Eric Dunford (ed769@georgetown.edu) McCourt School of Public Policy, Georgetown University

Learning Objectives

Asynchronous Materials

_

Visualizing Data

Relevant Slides

Grammar of Graphics

Relevant Slides

Code from the video

Continuous Data

Code from the video

Categorical Data

Code from the video

Many Plots

Code from the video

Customizing Plots

Code from the video

Practice

_

Question 1

_

Answer

Question 2

_

Answer

Question 3

_

Answer

Back to Course Website

Data Visualization

PPOL 670 | Introduction to Data Science

Lecture Materials for Week 5

Professor Eric Dunford (ed769@georgetown.edu)
McCourt School of Public Policy, Georgetown University