Spotify_Tidy_Tuesday

library(tidyverse)
library(lubridate)
library(patchwork)
library(ggrepel)
library(ggthemes)

spotify_songs <- read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-01-21/spotify_songs.csv')

arrow <- tibble(x1 = 25, x2 = 12, y1 = 320, y2 = 331.5)

spotify_plot <- spotify_songs %>%
  arrange(-duration_ms) %>%
  distinct(track_name, .keep_all = TRUE) %>%
  distinct(duration_ms, .keep_all = TRUE) %>%
  mutate(duration_s = round(duration_ms/1000)) %>%
  separate(col = track_album_release_date, into = c("Year", "Month", "Day"), sep = "-") %>%
  group_by(Year) %>%
  summarise(av_duration_s = mean(duration_s)) %>%
  ggplot(aes(x = Year, y = av_duration_s)) +
  geom_jitter(alpha = .5, size = 3) +
  coord_flip() +
  theme_economist() +
  theme(panel.grid.major = element_blank(), panel.grid.minor = element_blank()) +
  theme(panel.background = element_rect(fill = "white"))  +
  theme(plot.background = element_rect(fill = "white")) +
  scale_x_discrete(labels = c(1960, 1970, 1980, 1990, 2000, 2010, 2020), 
                     breaks = c(1960, 1970, 1980, 1990, 2000, 2010, 2020)) +
  theme(text = element_text(size = 12)) +
  labs(title = "",
       x = NULL, 
       y = "Average track duration (seconds)") +
  annotate("text", x = 30, y = 320,  size = 5, color = "gray20",
           label = "1971 and it all got a bit prog") +
  geom_curve(data = arrow, aes(x = x1, y = y1, xend = x2, yend = y2),
             arrow = arrow(length = unit(0.2, "cm")), size = 0.5,
             color = "gray20", curvature = 0.3)

  
plot_1971 <- spotify_songs %>%
  filter(playlist_genre == "rock") %>%
  arrange(-duration_ms) %>%
  distinct(track_name, .keep_all = TRUE) %>%
  distinct(duration_ms, .keep_all = TRUE) %>%
  mutate(duration_s = round(duration_ms/1000)) %>%
  separate(col = track_album_release_date, into = c("Year", "Month", "Day"), sep = "-") %>%
  filter(Year == 1971) %>%
  group_by(track_artist) %>%
  summarise(mean_duration_s = mean(duration_s)) %>%
  mutate(rank = rank(mean_duration_s)) %>%
  ggplot(aes(x = mean_duration_s, y = 0, label = track_artist, size = rank)) +
  geom_text_repel(force = 5, min.segment.length = 5) +
  labs(title = NULL,
       x = NULL,
       y = NULL) +
  theme_minimal() + 
  theme(axis.text.y = element_blank(), 
        axis.ticks.y = element_blank()) +
  scale_y_discrete(labels = c(-.02, 0, .02), breaks = c(-.02, 0, .02)) +
  scale_x_discrete(labels = c(200, 500), breaks = c(200, 500)) +
  guides(colour = FALSE) +
  theme(text = element_text(size = 12)) +
  annotate("text", x = 365, y = .35,  size = 12, 
           label = "Increasing average track duration in 1971 rock") +
  guides(size = FALSE)
  
spotify_plot / plot_1971