Context The Billboard 200 charts are a measure of the highest-selling (and in the era of streaming, most streamed) albums. This dataset was curated in the production of a piece on [Components][1]. While that piece deals with album length alone, I used Spotify's Python API to get the EchoNest acoustic features for as many songs from the Billboard 200 charts as possible. Content This databased contains two tables, "albums" and "acoustic_features": - "albums" contains 574,000 rows containing all albums in the Billboard 200 from 1/5/1963 to 1/19/2019. Each row contains the album’s place in the charts, the week of the chart, album name, artist name, and where available from Spotify and not null in the table, the number of tracks, and length of the album in milliseconds. - "acoustic_features" contains 340,000 rows containing acoustic data for tracks from Billboard 200 albums from 1/5/1963 to 1/19/2019. Each row contains track ID on Spotify, track name, album name, artist name, values for Spotify EchoNest acoustic data (acousticness, danceability, energy, instrumentalness, liveness, loudness, speechiness, key, time signature, and valence), duration in milliseconds, album ID on Spotify, and release date of the album. Contains no null values. [1]: https://components.one/posts/it-goes-on-album-length