We have worked with the Spotify API to analyse:
The most streamed songs of all time: Initial Dataset & Source.
The Top-100 songs of each year of the last 20 years (2000-2021): Initial Dataset & Source.
The most streamed artists: Initial Dataset & Source.
Through Spotify audio analysis API we have obtained information such as:
· Acousticness · Danceability · Duration · Energy · Instrumentalness · Liveness · Loudness · Speechiness · Tempo · Valence
You can find the Spotify Audio Analysis API Documentation and audio features definitions here.
Some of the questions we asked ourselves and have answered through data analysis are:
-
What similarities are there in the most popular music?
-
Is there any correlation between popularity and danceability / energy / tempo / etc?
-
From which year are the most streamed songs on spotify?
-
Do we like less danceable music or more danceable? (danceability)
-
Do we like vocal or instrumentalness music? (instrumentalness)
-
Does the music we like sound more positive or more negative? (valence)
-
What differences/similarities are there in the most popular music through time (from 2000 to 2021)?
-
How has the danceability, the tempo, the energy changed over the last few years?
-
Were there any of these variables that changed a significant amount?
-
Of the 1000 most popular artists, how many songs did they make on average?
-
How many of them became popular?
-
What is the most popular genre?
- You can find in the Final_SpotiData_Analysis.ipynb file all the analysis with pandas, numpy and matplotlib, as well as the information requests to the Spotify API.
- In Resources are the datasets.
- In Figures are the plots made with matplotlib.
- Also we added the Proposal and the Presentation.