PA 10: Exploring the Star Wars Universe

This task is complex. It requires many different types of abilities. Everyone will be good at some of these abilities but nobody will be good at all of them. In order to produce the best product possible, you will need to use the skills of each member of your group.

Goals for the Activity

• Apply methods of to use lists and iteration (using purrr) to extract data from various non-tabular data sets.
  
• Create new data sets through the cleaning, organization, and joining of data from various sources
  
• Create visualizations to explore the data
  
• May the force be with you!

THROUGHOUT THE Activity be sure to follow the Style Guide by doing the following:

• load the appropriate packages at the beginning of the Rmarkdown
  
• use proper spacing
  
• name all code chunks
  
• comment at least once in each code chunk to describe why you made your coding decisions
  
• add appropriate labels to all graphic axes

Review: Extracting Information from Different Data Sets

Here is information about the fist 7 Star Wars films:

View(sw_films) 

We are going to explore the data contained in several lists similar to this one (and the previously explored sw_people), combining skills from all of our previous R code learning experiences.

How do the following two codes compare?

sw_films[[4]][["title"]]
sw_films %>% pluck(4,"title")

Insert Answer Here

Suppose we want to pull out just the titles as a character vector, select the correct code (comment out the rest) to perform this action. You may want to run each line of code one at a time (remember Ctrl + Enter for Windows with your cursor on that line of code).

#comment out the incorrect codes
sw_films %>% map("title")
sw_films %>% map_chr("title")
sw_films %>% map_dfc("title")

Suppose we want to apply a function to count the number of specific kinds of ships and vehicles in our data

Notice that for each film, the “starships” vector contains links to information on those starships (though note this data is out of date and should be linked at swapi.dev, not swapi.co).

sw_films[[1]][["starships"]]

So if we can count the number of webpage links that would tell us the number of starships that appear in that movie. Here are three different ways to count the number of urls under starships. Can you think of another? (it is ok if you can’t). Compare and contrast how the three codes work differently to do the same thing.

sw_films %>% map(., "starships") %>% map_dbl(~length(.))
map_dbl(sw_films, ~length(.x$starships))
sw_films %>% map_dbl(., ~length(.x$starships))

Insert Answer

Part 1: Evaluating Hyperdrive in the Star Wars Episodes

We will use the third method from the previous section to extract out the information we want from sw_films. For each row, specify if we should use a regular map(), map_dbl(), or map_chr().

NOTE Sometimes code like this gets a little finicky in R if you try to run it with Ctrl + Enter. Instead, use the code chunk green arrow to run the whole code chunk or highlight all of the code and then use the shortcut to run it.

sw_ships_1 <- sw_films %>% {
  tibble(
    title = map____(., "title"), #character
    episode = map____(., "episode_id"), #numeric
    starships = map____(., ~length(.x$starships)), #numeric
    vehicles = map____(., ~length(.x$vehicles)), #numeric
    planets = map____(., ~length(.x$planets)) #numeric
  )}
sw_ships_1

Let’s do a bit more data cleaning to 1) assign the Trilogy classification to each episode, 2) calculate the total number of starships (which have hyperdrive) and vehicles (which do not have hyperdrive), and 3) calculate the proportion of total ships that have hyperdrive. Fill in the missing codes.

sw_ships <- sw_ships_1 ____  
  #create a new variable called trilogy
  ____(trilogy = case_when(episode %in% 1:3 ~ trilogies[1],
                             episode %in% 4:6 ~ trilogies[2],
                             episode %in% 7 ~ trilogies[3])) %>% 
  #create a new variable called total_ships which adds vehicles and starships together
  ____(total_ships = vehicles + starships) %>%  
  #create a new variable called prop that calculate the percent hyperdrive
  ____(prop = starships / total_ships * 100) 

Hyperdrive Use Across Films

Now, let’s make a plot examining how often hyperdrive ships appear in each episode. You will modify the code below to replicate the following visualization:

Fill in the blanks withe appropriate functions.

sw_ships %>% 
  #be sure to order titles by order/episode
  ggplot(aes(y = ____(title, desc(episode)), 
             x = prop)) + 
  #we want bars but our data is already summarized!
  geom_____(aes(fill = trilogy)) + 
  labs(
    title = "The Rise of Hyperdrive",
    subtitle = "Percentage of Ships with Hyperdrive Capability"
  ) +
  #you may need to install `scales` package if you haven't already
  scale_x_continuous(labels = scales::label_percent(scale = 1)) +
  theme_minimal() +
  #what aesthetic do we modify to change the bar color
  scale______viridis_d(end = 0.8) +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        legend.position = "bottom",
        legend.title = element_blank())

Canvas Quiz Question 1

Which movie has the second highest percentage of Hyperdrive ships?

Insert Answer Here

Hyperdrive Prevalence within the Universe

We can also look at a plot to see if there is a correlation between the total number of ships and the number with hyperdrive (starships).

Fill in the blanks withe appropriate functions.

sw_ships %>% 
  ggplot(aes(x = total_ships, 
             y = starships)) +
  #make points
  geom_____(aes(color = trilogy)) +
  #fit a model
  geom_____(method = "lm") +
  #what does geom_text() do?
  geom_text(aes(label = title), 
            vjust = -1, 
            hjust = "inward", 
            size = 2.75) +
  labs(title = "Hyperdrive Correlations",
       subtitle = "The Number of Ships with Hyperdrive vs Total Ships") +
  theme_minimal() +
  #what aesthetic do we want to modify the color of points?
  scale______viridis_d(end = 0.8) +
  theme(axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        legend.position = "bottom",
        legend.title = element_blank()) 

Canvas Quiz Question 2

What do you notice about the use of hyperdrive type vehicles in the episodes?

Insert Answer

Part 2: The Physical Features of Star Wars Characters

Recall the data for “people” in Star Wars:

View(sw_people)

We want to extract out name, height, and mass as character vectors (for now, we have to deal with some issues in height and weight later to change them into double type vectors) and keep films as a list for now. Fill in the correct map type functions for each one.

sw_peeps <- sw_people %>%  {
  tibble(
    name = ____(., "name"),  #character
    height = ____(., "height"), #character
    mass = ____(., "mass"), #character
    films = map____(., "films") #list
  )}
sw_peeps

Notice that the films column contains lists of urls for each film reference. Let’s pull out that same information from the sw_films data to have the title of the episode and the url as a character vector, and the episode number as a numeric value. Fill in the correct map type functions.

film_names <- sw_films %>% {
  tibble(
    episode_id = ____(., "episode_id"), #double
    episode_name = ____(., "title"), #character
    url = ____(., "url") #character
  )}
film_names

Now we can finish cleaning up our data by doing the following:

1. turn height and mass into numeric vectors;
   
2. match the films/urls to their episode_names and assign that back to sw_peeps.

sw_peeps2 <- sw_peeps %>% 
  #use a function from readr to extract the numbers and replace "unknown" with na
  mutate(height = parse_____(height, na = "unknown"),
         mass = parse_____(mass, na = "unknown")) %>%
  #unnest the lists in films
  ____(cols = c("films")) %>% 
  #join the film data with episodes names to the people data
  _____join(film_names, by = c("films" = "url")) %>% 
  #remove the `films` url from the data frame
  ____(-films) %>% 
  #add the variable trilogy
  ____(trilogy = case_when(episode_id %in% 1:3 ~ trilogies[1],
                           episode_id %in% 4:6 ~ trilogies[2],
                           episode_id %in% 7   ~ trilogies[3]))
sw_peeps2

Size of Characters in the Star Wars Universe

We can now create a plot of height and mass by trilogy group to see if the physique of characters differed across Trilogies (keeping in mind the third set of Trilogies is incomplete in this data set).

sw_peeps2 %>% 
  filter(name != "Jabba Desilijic Tiure") %>% #major outlier removed
  #map the correct aesthetics
  ggplot(aes(x = ____, 
             y = ____, 
             color = ____))+
  geom_point(position = "jitter") +
  geom_smooth(method = "lm", se = FALSE) +
  labs(x = "Height of Character (cm)",
       y = "Mass of Character (kg)",
       color = "Trilogy Group",
       title = "Character Characteristics in Star Wars") +
  theme_minimal() +
  scale_color_viridis_d(end = 0.8) 

Canvas Quiz Question 3

Write some code to identify who is is the heaviest (look at the graph to help guide this) Star Wars character (excluding Jabba Desilijic Tiure).

Insert Answer Here

OPTIONAL CHALLENGE PROBLEM

Your professor wants to use purrr to try and generate a height and mass scatterplot for each episode, but I don’t want to type out all that code. Here is where I got so far, but I am not convince this is the most sophisticated or effective way to do this. Do some research and see if you can find a way to put this process into production!

plots_sw <- sw_peeps %>% 
  nest(data = !episode_name) %>% 
  mutate(plot = map(data, ~ggplot(., aes(y=mass, x=height)) + 
                      geom_point() + 
                      geom_smooth(method = "lm", se = FALSE) + 
                      labs(title = paste0(episode_name))))

print(plots_sw$plot)