creator. writer. bread lover. and everything in between.

radarnonegif.gif

The Evolution of Dance Music

The Evolution of Dance Music

 

Visualizing trends in dance music from 1920s-present using Spotify’s API and R.

“Fred Astaire & Ginger Rogers Dancing” by Jordan Kay, Dribbble

“Fred Astaire & Ginger Rogers Dancing” by Jordan Kay, Dribbble

If I could time travel, I would go back to the 1940s and 1950s and dance my heart out. The end of World War II and rise of rock-n-roll ensured everyone had a pair of happy feet. Jazz and swing dances like the jitterbug to the Lindy hop were followed closely by the bop and twist, popularized by "Grease", "Hairspray", and Elvis' trigger-happy hips. 

dance_decade.png

My enthusiasm for casual dancing (e.g. at clubs, in-between studying, while microwaving food, anytime really) is rivaled by nothing other than tear-jerkingly beautiful data visualization. While I'm pretty average at both, I decided to analyze which attributes made a song more "danceable" and create a playlist of the best dance hits of each decade using Spotify's API, the spotifyr R package, and a list of 1300+ hits from 1920 to today. If you're just here for the dope playlist, feel free to jump to the bottom. Otherwise, enjoy the read!

What is danceability?

Spotify defines danceability as an attribute that "measures how suitable a song is for dancing based on a combination of musical elements including tempo, rhythm stability, beat strength, and overall regularity". For the purposes of this study, I used Spotify's definitions and values for different song attributes. Each attribute, excluding tempo, is valued on a scale of 0 to 1, with 1 being the most intense version of the attribute. 

Spotify Song Attributes
Attribute Description
Acousticness Measures how acoustic a song is.
Danceability Measures how suitable a song is for dancing based on a combination of musical elements including tempo, rhythm stability, beat strength, and overall regularity.
Energy Measures the intensity and activity of a song. Typically, energetic songs feel fast, loud, and noisy. Perceptual features contributing to this attribute include dynamic range, perceived loudness, timbre, onset rate, and general entropy.
Happiness Measures the musical positiveness conveyed by a song.
Popularity Measures how frequently played a song is on Spotify.
Speechiness Measures the presence of spoken words in a song.
Tempo The overall estimated speed or pace of a song in beats per minute (BPM).

How have dance hits changed over the last century?

radarnonegif.gif

Put away the banjo and break out the dubstep - people today dance to energetic, happier music. While speechiness and tempo have remained constant over the last century, energy and happiness have increased significantly.

What kind of songs did people enjoy dancing to?

Some attributes that make a song more dance-worthy are timeless. Generally, the happier and more energetic a song is, the more likely it will be a dance hit. However, tempo and speechiness display a diminishing returns relationship. The best dance songs rest at a sweet spot of 80-130 BPM. Anything slower or faster both drastically reduce how danceable a song is. In the same way, songs with a moderate amount of words make the best dance songs.

Interestingly, acousticness affects danceability differently across the decades. Up until the 1960s, a more acoustic song was generally perceived as more danceable, dropping off at fully acoustic songs. However, post-1970s, the trend flipped: acousticness made a song less danceable. 

Takeaways

This study was conducted primarily to explore the evolution of dance music over the last century. Current trends in what makes a song more danceable than others can help DJs, artists, and producers create hits based on quantifiable information like tempo or level of speechiness. Further probing of other song metrics' effects on danceability would provide deeper analysis. For example, using the Genius API and geniusr R package to analyze lyrics of dance hits or a musical dissection of the timing of beats, bass drops, and melodies could detail line-by-line preferences for making dance music. Beyond data, we can also compare these trends against different types of dancing (e.g. did older styles of dancing fit better with acoustic, lower-energy songs?) and historical context (e.g. could the drop in happiness levels in dance hits in the 1930s be attributed to the Great Depression?).

Overall, if this study and/or playlist inspired you to jump up and shake those hips, I've done my job. Go ahead - press play and dance your heart out, MJ. 

 

Top 10 songs by "danceability" - Spotify's internal measure of how danceable a song is - of each decade.

 
Best Dance Hits of Each Decade
Decade Song Artist Danceability Energy Speechiness Acousticness Happiness Tempo
1920s I Married the Bootlegger's Daughter Frank Crumit 0.91 0.25 0.12 1.00 0.67 118
1920s If You Knew Susie (like I Know Susie) Eddie Cantor 0.88 0.37 0.14 0.99 0.50 123
1920s My Wife Is On A Diet Leslie Sarony 0.83 0.65 0.06 1.00 0.87 124
1920s Kitchen Man Bessie Smith 0.81 0.08 0.06 0.99 0.48 116
1920s Happy Days Are Here Again Johnny Marvin 0.79 0.18 0.06 1.00 0.89 115
1920s Five Foot Two Eyes of Blue Art Landry 0.78 0.22 0.09 0.99 0.81 110
1920s I Want a Girl Dan Hornsby Trio 0.78 0.48 0.07 0.88 0.80 119
1920s A Good Man Is Hard to Find Bessie Smith 0.78 0.03 0.06 0.99 0.41 84
1920s Send Me to the 'Lectric Chair Bessie Smith 0.77 0.09 0.06 0.99 0.60 88
1920s Puttin' On The Ritz Harry Richman 0.80 0.35 0.11 0.94 0.84 104
1930s Puttin' On The Ritz Fred Astaire 0.85 0.35 0.16 0.99 0.79 106
1930s Reefer Man Cab Calloway 0.81 0.41 0.10 0.85 0.78 128
1930s Anything Goes Cole Porter 0.80 0.23 0.09 0.96 0.56 106
1930s Painting the Clouds with Sunshine Jack Hylton & His Orchestra 0.80 0.23 0.04 0.98 0.89 95
1930s Kickin' The Gong Around Cab Calloway 0.78 0.17 0.06 0.95 0.67 110
1930s Keep Young And Beautiful Harry Roy 0.77 0.59 0.12 0.99 0.85 103
1930s Dope Head Blues Victoria Spivey 0.77 0.15 0.04 0.98 0.52 78
1930s Nice Work If You Can Get It Billie Holiday 0.80 0.20 0.12 0.87 0.73 80
1930s The Saga of Jenny Gertrude Lawrence 0.76 0.08 0.09 0.98 0.63 114
1940s Walking The Floor Over You Ernest Tubb 0.88 0.32 0.06 0.93 0.82 115
1940s Pistol Packin' Mama Al Dexter & His Troopers 0.94 0.34 0.07 0.94 0.98 103
1940s Deep In the Heart of Texas Gene Autry 0.86 0.31 0.06 0.97 0.96 115
1940s Song Of The Volga Boatmen Glenn Miller 0.84 0.21 0.08 0.87 0.58 125
1940s It's A Good Day Peggy Lee 0.84 0.20 0.09 0.78 0.82 123
1940s Rum And Coca Cola The Andrews Sisters 0.84 0.24 0.11 0.46 0.69 80
1940s Doin' What Comes Naturally Ethel Merman 0.83 0.41 0.04 0.74 0.69 93
1940s Smoke! Smoke! Smoke! (That Cigarette) Tex Williams 0.83 0.55 0.26 0.72 0.96 128
1940s Lazy River The Mills Brothers 0.83 0.05 0.07 0.98 0.43 89
1950s Fever Peggy Lee 0.80 0.15 0.13 0.13 0.39 137
1950s Oh Oh I'm Falling In Love Again Jimmie Rodgers 0.80 0.44 0.06 0.87 0.71 95
1950s Seven Little Girls Sitting in the Back Seat The Avons 0.80 0.32 0.06 0.71 0.86 134
1950s Don't Let The Stars Get In Your Eyes Perry Como 0.80 0.42 0.04 0.81 0.96 122
1950s Butterfly Andy Williams 0.79 0.60 0.07 0.43 0.86 127
1950s I Get So Lonely The Four Knights 0.79 0.33 0.05 0.63 0.97 112
1950s Rocking Pneumonia and the Boogie Woogie Flu Huey "Piano" Smith 0.84 0.66 0.04 0.55 0.97 118
1950s A Guy Is A Guy Doris Day 0.84 0.22 0.08 0.90 0.88 122
1950s Heartbreak Hotel Elvis Presley 0.52 0.39 0.07 0.88 0.83 95
1950s Lollipop The Chordettes 0.68 0.32 0.09 0.79 0.73 144
1950s Johnny B. Goode Chuck Berry 0.51 0.69 0.08 0.73 0.97 168
1950s Jailhouse Rock Elvis Presley 0.49 0.86 0.07 0.17 0.94 84
1960s Green Onions Booker T. & the M.G.'s 0.79 0.38 0.03 0.81 0.89 138
1960s Sugar, Sugar The Archies 0.76 0.65 0.03 0.76 0.97 123
1960s Under The Boardwalk The Drifters 0.75 0.23 0.04 0.91 0.78 124
1960s Comment te dire adieu - It Hurts To Say Goodbye Françoise Hardy 0.73 0.45 0.06 0.20 0.84 124
1960s Runaway Del Shannon 0.71 0.82 0.09 0.09 0.74 144
1960s Footsteps In the Dark - Pt. 1 & 2 The Isley Brothers 0.69 0.41 0.05 0.31 0.86 81
1960s Sunshine Of Your Love Cream 0.68 0.49 0.04 0.44 0.83 115
1960s Mr. Soul Buffalo Springfield 0.67 0.55 0.03 0.10 0.96 135
1960s Monster Mash Bobby "Boris" Pickett 0.66 0.54 0.24 0.11 0.59 141
1960s Build Me Up Buttercup The Foundations 0.66 0.52 0.05 0.16 0.68 133
1970s Le Freak CHIC 0.87 0.75 0.05 0.05 0.93 119
1970s Boogie Oogie Oogie A Taste Of Honey 0.84 0.52 0.09 0.10 0.91 124
1970s Play That Funky Music Wild Cherry 0.82 0.87 0.06 0.12 0.90 109
1970s Grease Frankie Valli 0.82 0.37 0.09 0.26 0.87 110
1970s 50 Ways to Leave Your Lover Paul Simon 0.82 0.38 0.06 0.19 0.34 102
1970s Hot Stuff Donna Summer 0.81 0.73 0.03 0.00 0.96 120
1970s Bad Girls Donna Summer 0.80 0.68 0.04 0.01 0.96 121
1970s Dancing Machine The Jackson 5 0.79 0.79 0.09 0.12 0.85 109
1970s Knock Three Times Dawn 0.78 0.47 0.08 0.22 0.91 122
1970s Come and Get Your Love Redbone 0.78 0.81 0.03 0.03 0.97 105
1970s Y.M.C.A Village People 0.69 0.97 0.16 0.03 0.73 127
1970s Crocodile Rock Elton John 0.63 0.67 0.04 0.08 0.97 150
1970s Don't Go Breaking My Heart Elton John 0.74 0.86 0.04 0.20 0.78 131
1980s Billie Jean Michael Jackson 0.92 0.65 0.04 0.02 0.85 117
1980s Faith George Michael 0.89 0.48 0.12 0.01 0.61 96
1980s The Way You Make Me Feel Michael Jackson 0.88 0.85 0.15 0.05 0.54 114
1980s Jump (For My Love) The Pointer Sisters 0.86 0.71 0.09 0.11 0.85 134
1980s Wishing Well Terence Trent D'Arby 0.85 0.65 0.06 0.08 0.80 105
1980s What's Love Got to Do with It Tina Turner 0.85 0.41 0.07 0.22 0.79 98
1980s Wham Rap! (Enjoy What You Do?) Wham! 0.83 0.94 0.11 0.50 0.69 118
1980s How Will I Know Whitney Houston 0.83 0.54 0.04 0.20 0.93 119
1980s Eye In The Sky The Alan Parsons Project 0.82 0.42 0.03 0.56 0.52 112
1980s Eye of the Tiger Survivor 0.82 0.44 0.03 0.22 0.55 109
1980s I Wanna Dance with Somebody (Who Loves Me) Whitney Houston 0.71 0.82 0.05 0.21 0.87 119
1990s ATLiens OutKast 0.92 0.73 0.27 0.03 0.61 97
1990s Insane in the Brain Cypress Hill 0.90 0.61 0.21 0.03 0.77 102
1990s Still Not a Player Big Pun 0.89 0.70 0.28 0.23 0.90 94
1990s Gettin' Jiggy Wit It Will Smith 0.88 0.87 0.14 0.02 0.86 108
1990s No Diggity Blackstreet 0.87 0.65 0.29 0.33 0.65 89
1990s Jump Around House Of Pain 0.85 0.71 0.08 0.01 0.82 107
1990s Boombastic Shaggy 0.85 0.79 0.25 0.24 0.80 159
1990s Can I Kick It? A Tribe Called Quest 0.85 0.67 0.27 0.17 0.74 97
1990s The Rain (Supa Dupa Fly) Missy Elliott 0.82 0.58 0.18 0.12 0.73 160
1990s Summer Girls LFO 0.81 0.81 0.16 0.16 0.74 100
1990s This Is How We Do It Montell Jordan 0.79 0.70 0.06 0.02 0.78 104
2000s SexyBack Justin Timberlake 0.97 0.58 0.08 0.06 0.96 117
2000s One, Two Step Ciara 0.94 0.51 0.16 0.04 0.84 113
2000s Hollaback Girl Gwen Stefani 0.93 0.92 0.09 0.35 0.90 110
2000s Yeah! Usher 0.89 0.79 0.11 0.02 0.58 105
2000s Rock Your Body Justin Timberlake 0.89 0.71 0.14 0.20 0.82 101
2000s The Way You Move OutKast 0.87 0.60 0.05 0.13 0.64 126
2000s Ms. Jackson OutKast 0.84 0.81 0.27 0.14 0.62 95
2000s Work Out J. Cole 0.83 0.43 0.11 0.05 0.22 93
2000s Caught Up Usher 0.82 0.81 0.05 0.03 0.75 110
2000s Get The Party Started P!nk 0.81 0.90 0.05 0.00 0.96 129
2000s Hips Don't Lie Shakira 0.78 0.82 0.07 0.28 0.76 100
2010s Fancy Iggy Azalea 0.91 0.72 0.07 0.08 0.39 95
2010s Hotline Bling Drake 0.90 0.62 0.06 0.00 0.54 135
2010s Uptown Funk Mark Ronson 0.86 0.61 0.08 0.01 0.93 115
2010s Shape of You Ed Sheeran 0.83 0.65 0.08 0.58 0.93 96
2010s Work from Home Fifth Harmony 0.80 0.63 0.04 0.10 0.66 105
2010s Get Lucky Daft Punk 0.79 0.81 0.04 0.04 0.86 116
2010s One Dance Drake 0.79 0.62 0.05 0.01 0.37 104
2010s Only Girl (In The World) Rihanna 0.79 0.71 0.04 0.10 0.71 126
2010s Call Me Maybe Carly Rae Jepsen 0.78 0.60 0.04 0.02 0.61 120

How to recreate this project

Install and load the spotifyr, dplyr, tidyr, data.table, ggplot2, and fmsb data packages.

install.packages("spotifyr")
install.packages("dplyr")
install.packages("tidyr")
install.packages("data.table")
install.packages("ggplot2")
install.packages("fmsb")

Get API access from Spotify Developer and load into RStudio.

spotifyclient = "clientcode"
spotifykey = "keycode"

Sys.setenv(SPOTIFY_CLIENT_ID = spotifyclient)
Sys.setenv(SPOTIFY_CLIENT_SECRET = spotifykey)

access_token = get_spotify_access_token(client_id = Sys.getenv('SPOTIFY_CLIENT_ID'), client_secret = Sys.getenv('SPOTIFY_CLIENT_SECRET'))

Follow  playlists on Spotify that each contain top 100 songs of a decade using a Spotify account. Get playlist data using the get_user_audio_features function from spotifyr. The playlist urls are displayed below. 

# get playlist data from your user account
all_years = get_user_audio_features('username')

# rename all playlist urls as its decade
"1920s" = all_years$playlist_uri[all_years$playlist_uri == "7olpTqGzGrYuI76ZBlCLgs"]
"1930s" = all_years$playlist_uri[all_years$playlist_uri == "6hxNMGb4qtRZZbn02TqCFv"]
"1940s" = all_years$playlist_uri[all_years$playlist_uri == "5NOaxYV0zp0wpnWDilmO6K"]
"1950s" = all_years$playlist_uri[all_years$playlist_uri == "5ZNKySqQVGPXlIomBnmt0r"]
"1960s" = all_years$playlist_uri[all_years$playlist_uri == "2NFOUmp2wyR5CrXtKDkUkB"]
"1970s" = all_years$playlist_uri[all_years$playlist_uri == "6d16009NUz5vn0PCRbbMtN"]
"1980s" = all_years$playlist_uri[all_years$playlist_uri == "0l92UabLyRHluwgys6k2fC"]
"1990s" = all_years$playlist_uri[all_years$playlist_uri == "0rZJqZmX61rQ4xMkmEWQar"]
"2000s" = all_years$playlist_uri[all_years$playlist_uri == "2f6tXtN0XesjONxicAzMIw"]
"2010s" = all_years$playlist_uri[all_years$playlist_uri == "1tPWTwuxOLsE2Do1JQSUxA"]

After removing unused columns and cleaning up column names, separate the master dataframe into multiple dataframes separated by decade.

# create new dataframe named "df[decade]" based on "Decade" column
s <- setNames(split(all_years, with(all_years, interaction(Decade)), drop = TRUE), paste0("df", unique(all_years$Decade)))

# put new dataframes in global environment
list2env(s, globalenv())

Create bar chart of "Danceability" across the decades.

ggplot(all_years, aes(Decade, Danceability)) + 
  geom_bar(stat="identity", width=.5, fill=rgb(55,105,214, maxColorValue = 255)) + 
  theme(axis.text.x = element_text(angle=65, vjust=0.6)) +
  ggtitle("Which Decade Had the Most Danceable Hits?")

Create polar/radar charts for each decade's top hits song attributes.

# create new dataframe with average attributes of all songs in decade
polar1920s = c(mean(df1920$Danceability), mean(df1920$Energy), mean(df1920$Speechiness), mean(df1920$Acousticness), mean(df1920$Happiness), mean(df1920$Tempo))
polar1920s <- as.data.frame(t(polar1920s))

# rename columns
colnames(polar1920s) = c("Danceability", "Energy", "Speechiness", "Acousticness", "Happiness", "Tempo")

# create new rows for min and max of column values as the first two rows in dataframe
polar1920s = rbind(c(0, 0, 0, 0, 0, 0), polar1920s)
polar1920s = rbind(c(1, 1, 1, 1, 1, 200), polar1920s)

# create polar/radar chart
radarchart(polar1920s, axistype=1 , 
           pcol=rgb(243,94,90,0, maxColorValue = 255), pfcol=rgb(243,94,90,130, maxColorValue = 255), plwd=4 , 
           cglcol="grey", cglty=1, axislabcol="grey", caxislabels=seq(0,1,.25), cglwd=0.8,
           vlcex=0.8, title = "1920s")

Create line plots of "Danceability" by [other attribute].

ggplot(all_years, aes(otherattribute, Danceability, color = Decade)) +
  theme_minimal() +
  geom_smooth(se=FALSE)
 
 
Profilepic_squarespace.jpg