library(tidyverse) #data management and ggplot
library(igraph) # bipartite networks
library(RColorBrewer)
simpsons <- readr::read_delim("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-08-27/simpsons-guests.csv", delim = "|", quote = "")
simpsons## # A tibble: 1,386 x 6
## season number production_code episode_title guest_star role
## <chr> <chr> <chr> <chr> <chr> <chr>
## 1 1 002–102 7G02 Bart the Genius Marcia Wall~ Edna Kraba~
## 2 1 003–103 7G03 Homer's Odyssey Sam McMurray Worker
## 3 1 003–103 7G03 Homer's Odyssey Marcia Wall~ Edna Kraba~
## 4 1 006–106 7G06 Moaning Lisa Miriam Flynn Ms. Barr
## 5 1 006–106 7G06 Moaning Lisa Ron Taylor Bleeding G~
## 6 1 007–107 7G09 The Call of the~ Albert Broo~ Cowboy Bob
## 7 1 008–108 7G07 The Telltale He~ Marcia Wall~ Edna Kraba~
## 8 1 009–109 7G11 Life on the Fas~ Albert Broo~ Jacques
## 9 1 010–110 7G10 Homer's Night O~ Sam McMurray Gulliver D~
## 10 1 011–111 7G13 The Crepes of W~ Christian C~ Gendarme O~
## # ... with 1,376 more rows
- variable class description
- season integer Season of the show
- number character Episode number
- production_code character Production code for the episode
- episode_title character Episode Title
- guest_star character Guest star (actual name)
- role character Role in the show, either a character or themself
#how many roles have each guest star played?
simpsons %>%
group_by(guest_star) %>%
summarize(length(role))## # A tibble: 795 x 2
## guest_star `length(role)`
## <chr> <int>
## 1 'N Sync 1
## 2 "\"Weird Al\" Yankovic" 2
## 3 50 Cent 1
## 4 Aaron Paul 1
## 5 Adam Driver 1
## 6 Adam Savage 1
## 7 Adam Silver 1
## 8 Adam West 2
## 9 Aerosmith 1
## 10 Al Roker 1
## # ... with 785 more rows
# How many guest stars per season
simpsons %>%
group_by(as.numeric(season)) %>% #make season numeric so in proper numerical order
tally()## Warning: NAs introduced by coercion
## # A tibble: 31 x 2
## `as.numeric(season)` n
## <dbl> <int>
## 1 1 14
## 2 2 31
## 3 3 48
## 4 4 44
## 5 5 40
## 6 6 41
## 7 7 43
## 8 8 38
## 9 9 51
## 10 10 55
## # ... with 21 more rows
only picking last 5 seasons for ease of viewing connections
#select data to graph
guests<- simpsons %>%
filter(season %in% 26:30) %>%
select(guest_star, number)
#create contigency matrix
guest_mat<- as.data.frame.matrix(table(guests))
#graph incidence
guests_inc1<- graph.incidence(guest_mat)
# creat bipartite networks
guests_pr1<- bipartite.projection(guests_inc1)
#select just project1 from bipartite (vertex names= guest star names)
collab_guests<- guests_pr1$proj1
#crude plot, will need some serious formatting
plot(collab_guests)
#filter for degree of less than 5
less_deg<- which(degree(collab_guests)<5)
#pull out those who do not have many connections
collab_guests_deg5<- delete.vertices(collab_guests, less_deg)
#formatting for better visual
plot(collab_guests_deg5,
vertex.label=V(collab_guests_deg5)$name,
vertex.size=degree(collab_guests_deg5)/3,
vertex.color="darkred",
layout=layout_with_kk,
main="Simpsons Guest Starring Network (Seasons 26 through 30)",
vertex.label.cex = .60)
#look at total degree by guest
prom.df <- data.frame(
names= V(collab_guests_deg5)$name,
deg.guests = degree(collab_guests_deg5))
prom.df<- prom.df %>%
filter(deg.guests>8)
#bar charts_ top degree
p <- ggplot(prom.df, aes(x = reorder(names, +deg.guests), y = deg.guests)) +
geom_bar(stat = "identity", fill="Salmon") + theme_minimal()
p+ coord_flip() + ggtitle("Prominent Guest Stars in the Simpsons' Guest Star Network\n (Seasons 26-30)") + labs(x="Guest Star Names", y= "Number of Episodes shared with other Guests (Degree)")