• Scientific notation is annoying. How do you get rid of it?
  • options(scipen=999)
    
• Is there a shortcut for clearing the output in R Notebook?
  • I didn’t find one, but let me know if you do!
• What do the # values do in your code?
  • Use the '#' to comment code in R

• Yardstick for plotting data in R
  
• tidyverse package
  
• Virtually no limits to plot types
  
• Add-on packages available
• Plotting resources
  • ggplot2 Launch Page
  • ggplot2 Cheat Sheet
  • Top 50 ggplot2 Visualizations

1 discrete variable (plus other optional discrete and/or continuous variables)

1 continuous variable

2 continuous variables

1 continuous variable + date variable

candi %>% ggplot(aes(primary_general)) + geom_bar()
candi %>% ggplot() + geom_bar(aes(primary_general))

candi %>% ggplot(aes(primary_general, fill = primary_general)) + geom_bar()

candi %>% ggplot(aes(primary_general, fill = type)) + geom_bar()

• Build a bar chart with crime data that shows how many incidences occurred in sector
• Only include these sector values: c('B', 'E', 'D', 'R', 'O', 'C', 'K')
  
• Color the bars using crime_subcategory
  
• Only include these crime_subcategory values: c('ROBBERY-STREET', 'THEFT-BICYCLE', 'AGGRAVATED ASSAULT', 'TRESPASS')

Hint

crime %>%
  filter() %>%
  ggplot()

candi %>% 
  filter(! primary_general %in% NA) %>% 
  ggplot(aes(primary_general, fill = primary_general)) + 
  geom_bar() + 
  labs(x = 'Election Type', y = 'Donations (#)')

crime %>% 
  filter(
    sector %in% c('B', 'E', 'D', 'R', 'O', 'C', 'K') &
    crime_subcategory %in% c('ROBBERY-STREET', 'THEFT-BICYCLE'
                             , 'AGGRAVATED ASSAULT', 'TRESPASS')
    ) %>%
  ggplot(aes(sector, fill = crime_subcategory)) + geom_bar() + 
  labs(x = 'Election Type', y = 'Donations (#)', fill = 'Check out these colors')

Try removing the legend name from your sector plot

Try removing the legend name from your sector plot

crime %>% 
  filter(sector %in% c('B', 'E', 'D', 'R', 'O', 'C', 'K')) %>%
  ggplot(aes(sector, fill = crime_subcategory)) + geom_bar() + 
  labs(x = 'District Sectors', y = 'Incidences by Event\nClearance Group (#)'
       , fill = element_blank())

Chart titles and subtitles

• Consider making titles (or subtitles) descriptive
• Don’t over use subtitles

# Option 1
labs(title = NULL, subtitle = NULL)

# Option 2
ggtitle(title = NULL, subtitle = NULL)

candi %>% 
  ggplot(aes(primary_general, fill = primary_general)) + 
  geom_bar() + 
  labs(
    x = 'Election Type'
    , y = 'Donations (#)'
    , fill = element_blank()
    , title = 'Number of Donations by Election Type'
    , subtitle = 'Over 6,000 full election cycle donations'
    )

candi %>% 
  filter(! primary_general %in% NA) %>% 
  ggplot(aes(primary_general, fill = primary_general)) + 
  geom_bar() + 
  labs(
    x = 'Election Type'
    , y = 'Donations (#)'
    , fill = element_blank()
    , title = 'Number of Donations by Election Type'
    , subtitle = 'Over 6,000 full election cycle donations'
    ) + 
  scale_y_continuous(labels = comma) 

• Change layer and background colors
• Change fonts
• Change plot borders/boundaries and ticks
• Pre-built themes vs. custom themes
  • ggplot2 themes
  • ggthemes themes

install.packages('ggthemes')
library(ggthemes)

Selected ggplot2 themes

theme_classic()
theme_minimal()
theme_dark()

Selected ggthemes themes

theme_stata() + scale_colour_stata() # scale_fill_stata()
theme_economist() + scale_colour_economist() # scale_fill_economist()
theme_fivethirtyeight() + scale_color_fivethirtyeight() # scale_fill_fivethirtyeight()
theme_wsj() + scale_colour_wsj() # scale_fill_wsj()
theme_pander() + scale_colour_pander() # scale_fill_pander()
theme_hc(bgcolor = "darkunica") + scale_colour_hc("darkunica") # scale_fill_hc("darkunica")

candi %>% 
  filter(! primary_general %in% NA) %>% 
  ggplot(aes(primary_general, fill = primary_general)) + 
  geom_bar() + 
  labs(
    x = 'Election Type'
    , y = 'Donations (#)'
    , fill = element_blank()
    , title = 'Number of Donations by Election Type'
    , subtitle = 'Over 6,000 full election cycle donations'
    ) + 
  scale_y_continuous(labels = comma) +
  theme_economist() + 
  scale_fill_economist()

Economist theme

• You can create custom themes with theme()
• Greater control over chart aesthestics
• Use custom themes in place of pre-built themes if pre-built themes don’t cut it
• Also, use custom themes to augment pre-built themes

t <- theme(
  plot.title = element_text(size=14, face="bold", vjust=1)
  , plot.background = element_blank()
  , panel.grid.major = element_blank()
  , panel.grid.minor = element_blank()
  , panel.border = element_blank()
  , panel.background = element_blank()
  , axis.ticks = element_blank()
  , axis.text = element_text(colour="black", size=12)
  , axis.text.x = element_text(angle=45, hjust=1)
  , legend.title = element_blank()
  , legend.position = "none"
  , legend.text = element_text(size=12)
  )

• Count, sum, average, and identify max and min values
  
• Example questions you could answer
  • In candi, which type received the most money in campaign donations from 'NON PARTISAN'?
  • In candi, which five contributor_state values saw the largest number of cash donations, excluding 'WA' and NA values?
  • In candi, what is the largest average donation amount for contributor_state?
  • In crime, what how many crimes were reported (reported_year) one and two years after the year in which they occurred (occurred_year)?
  • In crime, of 'THEFT-ALL OTHER' values in crime_subcategory which primary_offense_description value has the fewest incidences?

Example
In candi, which type received the most money in donations from 'NON PARTISAN'?

candi %>% 
  group_by(type, party) %>% 
  summarise(dollars = sum(amount, na.rm = TRUE)) %>%
  filter(party %in% 'NON PARTISAN')
## # A tibble: 3 x 3
## # Groups:   type [3]
##   type                party         dollars
##   <chr>               <chr>           <dbl>
## 1 Candidate           NON PARTISAN 1803901.
## 2 Political Committee NON PARTISAN     170.
## 3 <NA>                NON PARTISAN   64202.

How would you turn the tabular output data into a chart?

## # A tibble: 3 x 3
## # Groups:   receipt_year [1]
##   receipt_year type                total_amount
##          <int> <chr>                      <dbl>
## 1         2015 Candidate                572504.
## 2         2015 Political Committee      883725.
## 3         2015 <NA>                      59190.

crime %>% 
  transmute(
    sector
    , year_dif = reported_year - occurred_year
  ) %>% 
  group_by(sector) %>% 
  summarise(
    avg_year_dif = mean(year_dif, na.rm = TRUE)
    , n = n()
  ) %>% 
  filter(n > 1000 & ! sector %in% NA) %>% 
  ggplot(aes(reorder(sector, -avg_year_dif), avg_year_dif)) +
  geom_bar(stat = 'identity') + 
  theme_classic()

Histograms: 1 continuous variable

candi %>% 
  filter(amount > 0 & amount < 1000) %>%
  ggplot(aes(amount)) + geom_histogram() # 'bins' is the argument you use to indicate the number of columns in your histogram

Histograms: 1 continuous variable

candi %>% 
  filter(amount > 0 & amount < 1000) %>%
  ggplot(aes(amount)) + geom_histogram(bins = 10) # 'bins' is the argument you use to indicate the number of columns in your histogram

Histograms: 1 continuous variable

candi %>% 
  filter(amount > 0 & amount < 1000) %>%
  ggplot(aes(amount)) + geom_density() # works like geom_histogram(); continuous curve

Scatterplots: 2 continuous variables

candi %>% 
  filter(amount > 0 & amount < 1000000) %>%
  ggplot(aes(receipt_date, amount)) + 
  geom_point()

Scatterplots: 2 continuous variables

candi %>% 
  ggplot(aes(receipt_date, amount)) + 
  geom_point()

geom_count() # adds a size component to your scatterplot
geom_jitter() # adjusts cartesian location of point relative to other points

 crime %>% 
  filter(! precinct %in% NA) %>%
  ggplot(aes(reported_date, occurred_date, color = precinct)) + 
  geom_point() + 
  theme_classic()

• Create a chart from the crime dataset
  
• Create a new variable with mutate() called occurred_time_ampm that tells whether an incident occurred in the AM or PM.
  
• Make sure your axes are labeled correctly
• Give your chart a title
• Use occurred_time_ampm in the color or fill arguments in aes()
• Use a pre-built theme

crime %>% 
  mutate(occurred_time_ampm = ifelse(occurred_time >= 1200, 'PM', 'AM')) %>% 
  filter(
    ! occurred_time_ampm %in% NA & 
    ! crime_subcategory %in% NA
    ) %>%
  group_by(crime_subcategory, occurred_time_ampm) %>%
  summarise(n = n()) %>%
  ungroup %>%
  group_by(crime_subcategory) %>%
  mutate(total = sum(n, na.rm = TRUE)) %>%
  filter(total >= 100) %>%
  ggplot(aes(reorder(crime_subcategory, n), n, fill = occurred_time_ampm)) + 
  geom_bar(stat = 'identity') + 
  coord_flip() + 
  theme_wsj() + 
  scale_colour_wsj() + 
  labs(title = 'Most incidence occur\nin the AM', fill = 'Time of Day')

• Create a chart from the candi dataset
  
• Make some comparison
• Give your chart a title
• Use a pre-built theme

candi %>% 
  filter(
    ! party %in% c(NA, 'OTHER', 'NONE') & 
    amount > 0 & 
    receipt_date >= '2008-01-01' & 
    receipt_date <= '2020-12-31'
    ) %>% 
  group_by(party, receipt_date) %>%
  summarise(amount = sum(amount)) %>%
  mutate(amount = cumsum(amount)) %>% # You can see that I use cumsum() in mutate()
  ggplot(aes(x = receipt_date, y = amount, color = party)) + 
  geom_line(size = 1, stat = 'identity') + 
  theme_wsj() + 
  scale_y_continuous(labels = dollar) +
  scale_colour_wsj() + 
  labs(
    title = 'Democrats receive nearly 1\nmillion more in donations than\nnext closest party',
    colour = element_blank()
    )

crime %>% 
  group_by(neighborhood, crime_subcategory) %>% 
  summarise(count = n()) %>%
  filter(
    neighborhood %in% c('DOWNTOWN COMMERCIAL', 'CAPITOL HILL', 'NORTHGATE', 'UNIVERSITY', 'BALLARD SOUTH', 'BELLTOWN') & 
    crime_subcategory %in% c('CAR PROWL', 'THEFT-ALL OTHER', 'THEFT-SHOPLIFT', 'BURGLARY-RESIDENTIAL', 'MOTOR VEHICLE THEFT')
    ) %>%
  ggplot(aes(x = neighborhood, y = crime_subcategory, fill = count)) +
  geom_tile() + 
  theme_classic() + 
  theme(axis.text.x = element_text(angle = 45, hjust = 1))