4 Lab III: Univariate Visualizations

## Packages
library(tidyverse)
library(ggthemes)

## Data Loading
## Replace this with your working directory
load("~/GOVT8001/Lab 3/white_minwage.RData")

This lab shows step-by-step how to build basic histograms and barplots with library(ggplot2)

4.1 Histograms with `library(ggplot2)`

Histograms are good to visualize the distribution of one continuous variable.

4.1.1 Step One

Specify the tibble to be piped into ggplot()

## Building A Basic Histogram 
df.county

4.1.2 Step Two

Pipe the tibble into ggplot()
Specify the variable of interest with ggplot(aes(x = X))

## Building A Basic Histogram 
df.county %>%
  ggplot(aes(x = minimum.wage))

4.1.3 Step 3

Use + instead of %>% to move to next line in ggplot()
geom_histogram() creates the histogram

## Building A Basic Histogram 
df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..))

4.1.4 Step 4

Customization of theme, colors, and labels.
You can also save the object above and customize it later as shown below
- Use col = and fill = in geom_histogram() to set colors

## Building A Basic Histogram 
df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..), col = "dark red", fill = "tomato")

Use + theme() to set the theme
- library(ggtheme) has themes from your favorite publications!

## Building A Basic Histogram 
df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..), col = "dark red", fill = "tomato") + 
  theme_minimal()

Use + labs to set labels
- title = for a title
- subtitle = for a subtitle
- x = for x axis label and y = for y axis label
- caption = for caption to include data source or note

## Building A Basic Histogram 
df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..), col = "dark red", fill = "tomato") + 
  theme_minimal() +
  labs(title = "Distribution of Minimum Wage", subtitle = "All US Counties 1996 - 2016",
       x = "Minimum Wage", caption = "Data Source: Markovich & White (2022)",
       y = "Density")

## Or You Can Save the Basic Plot and Experiment

p <- df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..), col = "dark red", fill = "tomato")
p

p + 
  theme_minimal() +
  labs(title = "Distribution of Minimum Wage", subtitle = "All US Counties 1996 - 2016",
       x = "Minimum Wage", caption = "Data Source: Markovich & White (2022)",
       y = "Density")

4.2 Barplots with `library(ggplot2)`

Barplots are good for visualizing distributions by groups. The steps here follow closely what we did for the histogram.

4.2.1 Step One

We will be using simulated data for this example.
First we need to format our simulated data into something we can use for the barplot with the skills we learned last week.

## Simulated Data
df <- data.frame("age" = c("18 to 29", "36 to 50", "51 to 64", "65+"),
                "popPct" = c(29, 21, 30, 20),
                "svyPct" = c(19, 21, 32, 28))
df

##        age popPct svyPct
## 1 18 to 29     29     19
## 2 36 to 50     21     21
## 3 51 to 64     30     32
## 4      65+     20     28

## Building A Basic Barplot
df %>% 
  rename(Population = popPct, Survey = svyPct) %>%
  pivot_longer(-age, names_to = "Group", values_to = "Percent")

## # A tibble: 8 × 3
##   age      Group      Percent
##   <chr>    <chr>        <dbl>
## 1 18 to 29 Population      29
## 2 18 to 29 Survey          19
## 3 36 to 50 Population      21
## 4 36 to 50 Survey          21
## 5 51 to 64 Population      30
## 6 51 to 64 Survey          32
## 7 65+      Population      20
## 8 65+      Survey          28

4.2.2 Step Two

Pipe the tibble into ggplot()
Specify the variable of interest with ggplot(aes(x = X))
Since we want to show the distribution of X by some group, we can use fill = to specify the group

## Building A Basic Barplot
df %>% 
  rename(Population = popPct, Survey = svyPct) %>%
  pivot_longer(-age, names_to = "Group", values_to = "Percent") %>%
  ggplot(aes(x = age, y = Percent, fill = Group))

4.2.3 Step 3

Use + instead of %>% to move to next line in ggplot()
geom_bar() creates a barplot

## Building A Basic Barplot
df %>% 
  rename(Population = popPct, Survey = svyPct) %>%
  pivot_longer(-age, names_to = "Group", values_to = "Percent") %>%
  ggplot(aes(x = age, y = Percent, fill = Group)) +
  geom_bar(stat = "identity", position = "dodge")

4.2.4 Step 4

Now, we can customize just like above with the histogram.
scale_fill_grey() changes the color palette to greyscale

## Building A Basic Barplot 
df %>% 
  rename(Population = popPct, Survey = svyPct) %>%
  pivot_longer(-age, names_to = "Group", values_to = "Percent") %>%
  ggplot(aes(x = age, y = Percent, fill = Group)) +
  geom_bar(stat = "identity", position = "dodge") +
  scale_fill_grey() + 
  theme_minimal() +
  labs(x = "Age Group", y = "Percent",
       title = "Population and Survey Sample Proportions by Age Group")

4.3 Now Make Your Own Histogram or Barplot!

df.county %>%
  ggplot(aes(x = minimum.wage)) + 
  geom_histogram(aes(y = ..density..), col = "pink", fill = "black") +
  theme_economist() +
  labs(title = "Our Beautiful Plot")