ggplot2

Quick Introduction

Berk Orbay

MEF BDA 503

What is dplyr?

ggplot2 is an R package specialized on data visualization. It is both a companion to the dplyr data manipulation package and a whole library by itself. Within a well defined framework, you can

easily draw a large number plot types with the same methodology
draw sophisticated plots
multiple well adjusted plots (e.g, 2x2)
do advanced theming and custom theming
parametric plots

Official page: https://ggplot2.tidyverse.org/

Why ggplot2?

ggplot2 has the same UX in mind with dplyr and is being developed by the same person (Hadley Wickham) and their team.

Based on Grammar of Graphics (hence the “gg”).
Start with a “canvas” (ggplot()) and apply data (aes(x,y,color)) and layers (geom_*) on it
Add labels, annotations, titles, themes easily.

Fundamentals

Important Full notes are on Book of EDA

##  If you have never installed ggplot2 before
##  install.packages("tidyverse")

##  First load the library
##  We will also use dplyr therefore tidyverse is better
library(tidyverse)

## We are going to use starwars datasets to show examples
starwars

# A tibble: 87 × 14
   name     height  mass hair_color skin_color eye_color birth_year sex   gender
   <chr>     <int> <dbl> <chr>      <chr>      <chr>          <dbl> <chr> <chr> 
 1 Luke Sk…    172    77 blond      fair       blue            19   male  mascu…
 2 C-3PO       167    75 <NA>       gold       yellow         112   none  mascu…
 3 R2-D2        96    32 <NA>       white, bl… red             33   none  mascu…
 4 Darth V…    202   136 none       white      yellow          41.9 male  mascu…
 5 Leia Or…    150    49 brown      light      brown           19   fema… femin…
 6 Owen La…    178   120 brown, gr… light      blue            52   male  mascu…
 7 Beru Wh…    165    75 brown      light      blue            47   fema… femin…
 8 R5-D4        97    32 <NA>       white, red red             NA   none  mascu…
 9 Biggs D…    183    84 black      light      brown           24   male  mascu…
10 Obi-Wan…    182    77 auburn, w… fair       blue-gray       57   male  mascu…
# ℹ 77 more rows
# ℹ 5 more variables: homeworld <chr>, species <chr>, films <list>,
#   vehicles <list>, starships <list>

Let’s recall starwars tibble from dplyr.

Scatter plot `geom_point`

select simply displays columns

ggplot(starwars %>% select(name, height, mass) %>% filter(mass < 1000) %>% filter(complete.cases(.)), aes(x = height, y = mass)) +
    geom_point()

We use + instead of the pipe operator (%>%) to connect layers.

Scatter plot (cont.) `geom_point`

Let’s add some categorization with color.

ggplot(starwars %>% select(name, species, height, mass) %>% filter(mass < 1000) %>% filter(complete.cases(.)), aes(x = height, y = mass)) +
    geom_point(aes(color = species))

Scatter plot (cont.) `geom_point`

Let’s also add some labels and themes.

ggplot(starwars %>% select(name, species, height, mass) %>% filter(mass < 1000) %>% filter(complete.cases(.)), aes(x = height, y = mass)) +
    geom_point(aes(color = species)) +
    labs(title = "Star Wars Characters by Height and Mass", subtitle = "Colors by species", x = "Height (in meters)", y = "Mass (in KG)", color = "Species") +
    theme_minimal() +
    theme(legend.position = "bottom")

Line plot `geom_line`

Let’s use another data for line plot, EUStockMarkets data from base R.

stock_df <- as_tibble(EuStockMarkets) %>% mutate(date = lubridate::as_date("1991-01-01") + 1:nrow(.), .before = everything())
print(stock_df)

# A tibble: 1,860 × 5
   date         DAX   SMI   CAC  FTSE
   <date>     <dbl> <dbl> <dbl> <dbl>
 1 1991-01-02 1629. 1678. 1773. 2444.
 2 1991-01-03 1614. 1688. 1750. 2460.
 3 1991-01-04 1607. 1679. 1718  2448.
 4 1991-01-05 1621. 1684. 1708. 2470.
 5 1991-01-06 1618. 1687. 1723. 2485.
 6 1991-01-07 1611. 1672. 1714. 2467.
 7 1991-01-08 1631. 1683. 1734. 2488.
 8 1991-01-09 1640. 1704. 1757. 2508.
 9 1991-01-10 1635. 1698. 1754  2510.
10 1991-01-11 1646. 1716. 1754. 2497.
# ℹ 1,850 more rows

Let’s also make it more “plotable”.

stock_df_long <- stock_df %>% pivot_longer(-date,names_to="symbol",values_to="close")
print(stock_df_long)

# A tibble: 7,440 × 3
   date       symbol close
   <date>     <chr>  <dbl>
 1 1991-01-02 DAX    1629.
 2 1991-01-02 SMI    1678.
 3 1991-01-02 CAC    1773.
 4 1991-01-02 FTSE   2444.
 5 1991-01-03 DAX    1614.
 6 1991-01-03 SMI    1688.
 7 1991-01-03 CAC    1750.
 8 1991-01-03 FTSE   2460.
 9 1991-01-04 DAX    1607.
10 1991-01-04 SMI    1679.
# ℹ 7,430 more rows

Line plot (cont.) `geom_line`

Let’s generate a plot to compare all four stock market indices.

ggplot(stock_df_long,aes(x=date,y=close,color=symbol)) + geom_line()

Line plot (cont.) `geom_line`

Let’s theme it up a bit with the dark theme.

ggplot(stock_df_long,aes(x=date,y=close,color=symbol)) + 
    geom_line() + 
    labs(x="Date",y="Index Level at Close",
    title="Comparison of Stock Market Indices",subtitle="Period between 1991-1996",color="Index") + 
    theme_dark()

Bar plot `geom_bar`

Back to starwars. Let’s see the frequency of “eye color” of characters and draw a bar plot.

eye_color_df <- starwars %>% count(eye_color)
eye_color_df

# A tibble: 15 × 2
   eye_color         n
   <chr>         <int>
 1 black            10
 2 blue             19
 3 blue-gray         1
 4 brown            21
 5 dark              1
 6 gold              1
 7 green, yellow     1
 8 hazel             3
 9 orange            8
10 pink              1
11 red               5
12 red, blue         1
13 unknown           3
14 white             1
15 yellow           11

ggplot(eye_color_df,aes(x=eye_color,y=n)) + geom_bar(stat="identity")

Bar plot (cont.) `geom_bar`

Let’s make it more detailed with whether the eye color belongs to a human or not and sort by frequency. Notice how we

eye_color_species_df <- starwars %>% mutate(is_human=(species == "Human" & !is.na(species))) %>% group_by(is_human) %>% count(eye_color) %>% ungroup()
ggplot(eye_color_species_df,
    aes(x=reorder(eye_color,-n,function(x){sum(x)}),y=n,fill=is_human)) + 
    geom_bar(stat="identity",position="stack") + labs(title="Star Wars Characters by Eye Color",x="Eye Color",y="Number of Characters",fill="is human?")

Pie chart `geom_bar` + `coord_polar`

Make some minor modifications (move x to fill) and add coord_polar; now you get a pie chart.

ggplot(eye_color_species_df,
    aes(x="",y=n,fill=reorder(eye_color,n,function(x){sum(x)}))) + 
    geom_bar(stat="identity",position="stack") + 
    labs(title="Star Wars Characters by Eye Color",x="Eye Color",y="Number of Characters",fill="is human?") + 
    coord_polar("y") + theme_minimal()

Advanced `ggplot2`

There is always more to learn. Here are some examples.

There are much more plot types see the cheat sheet
facet_grid for multiple plots (see).
scale_* functions for fine tuning labels, axes etc.
It is possible to add custom themes (see ggthemes library)

See extra resources section on Book of EDA.

Thanks!

Course webpage https://mef-bda503.github.io/.

ggplot2

What is dplyr?

Why ggplot2?

Fundamentals

Scatter plot geom_point

Scatter plot (cont.) geom_point

Scatter plot (cont.) geom_point

Line plot geom_line

Line plot (cont.) geom_line

Line plot (cont.) geom_line

Bar plot geom_bar

Bar plot (cont.) geom_bar

Pie chart geom_bar + coord_polar

Advanced ggplot2

Thanks!

Scatter plot `geom_point`

Scatter plot (cont.) `geom_point`

Scatter plot (cont.) `geom_point`

Line plot `geom_line`

Line plot (cont.) `geom_line`

Line plot (cont.) `geom_line`

Bar plot `geom_bar`

Bar plot (cont.) `geom_bar`

Pie chart `geom_bar` + `coord_polar`

Advanced `ggplot2`