Billboard Visualization with Tableau
I was recently reviewing few topics to visualize in Tableau and I also wanted to start off the new year with writing about something that interests me. So along came this post which is about a work in Tableau that explores the trends in data of the Billboard Hot 100 Singles since 1958 using the dataset by kcmillerean.
The dashboard above which is hosted in Tableau Public focuses on tracks’ performance record on Tableau categorized in four Tabs:
Overview
This is the first tab and shows the journey of the selected track through weeks, marking their highest and lowest rank in between. We can get to know the popularity of a song with the duration of stay in the charts as well as the rate of ascent and descent in ranks.
Track Features
The second tab is about the technical aspect of a song like tempo, loudness, etc. The metrics to create this view had to be brought from the second dataset using a join on field SongID. The bar chart is made to resemble an equalizer of an audio player and shows the different metrics on its bars.
Week Top 10
The third tab describes the pattern of tracks taking up the top 10 weekly ranks in a selected month of any year that again gives some curious stories behind different songs filling these ranks.
Year Top 10
The final tab is the more innovative than other and explores the genre split of the top 100 songs in a year. So instead of selecting a drop down, clicking on different small red stars on the chords of the guitar image would filter off the year 100 table below. The technique used is called background image in a view as outlined in this kb article. All throughout a year, different tracks enter and exit a particular rank on weekly basis. So a track stays in different ranks for specific amount of time before finally exiting from the charts. If we pick tracks to have stayed for the maximum duration at various positions starting from 1 and down till 100, we would have a list of 100 most popular year end songs and this is the logic used for creating the year 100 table.
Information Ribbon
Just below the tabs on top, sits two data driven insights. One of them tells us the artist with maximum number of songs featured in Billboard charts using nested LOD expression to get the highest value as described in this KB article.
The other insight also uses the same technique but gets the song that stayed for the longest duration on billboard charts.
This is still a work in progress and may have further changes in future about which I will make edits in the post.
