In IAjimi/idk: Spotify Top 200 Tracks

knitr::opts_chunk$set(echo = TRUE)

library("ggplot2")
library("dplyr")
library("idk")

Chart Position

This Vignette briefly shows some of the information one can derive from the chart_position dataset.

First, we can look at streams per artist over time.

# Total Streams per Artist per Day
time_position <- chart_position %>% 
  group_by(Artist, Date) %>%
  summarize(Streams = sum(Streams),
            per = Streams / sum(.$Streams)) %>%
  arrange(desc(Date))

# Find Top Artists
top_artists <- time_position %>% arrange(desc(Streams)) %>% distinct(Artist)

# Graph: Total Streams per Date per Artist
ggplot(time_position, aes(Date, Streams, group = Artist)) +
  geom_line(color = "grey") +
  geom_line(data = filter(time_position, Artist %in% top_artists$Artist[1:5]) , aes(color = Artist)) +
  labs(title = "Streams Per Artist Over Time")

We can also who were the most streamed artists from 2017 to 2019 and in 2019 alone:

## Most Streamed Artists Overall
chart_position %>% 
  group_by(Artist) %>%
  summarize(Streams = sum(Streams),
            per = Streams / sum(.$Streams)) %>%
  arrange(desc(Streams))

## Most Streamed Artists in 2019
chart_position %>% 
  filter(Date >= "2019-01-01") %>%
  group_by(Artist) %>%
  summarize(Streams = sum(Streams),
            per = Streams / sum(.$Streams)) %>%
  arrange(desc(Streams))

We can look at which songs were the most streamed throughout this time period:

## Most Streamed Songs Overall
chart_position %>% 
  group_by(Artist, Track) %>%
  summarize(Streams = sum(Streams),
            per = Streams / sum(.$Streams)) %>%
  arrange(desc(Streams))

Finally, we can find the most streamed songs per day, as well as the songs which were in the top position the longest:

## Most Streamed Songs in A Day
chart_position %>% arrange(desc(Streams)) %>% select(Track, Artist, Date, Streams)

## Songs Longest in Number 1 Position
chart_position %>% filter(Position == 1) %>% group_by(Track, Artist) %>% count() %>% arrange(desc(n))