Also, a lesson in Altair python visualisations
This is an ancient technique. It’s called a power stance. You own the universe, so raise your goblet of rock; it’s a toast to those who rock. Now smile and nod your head and let me see your eyeballs wide like there’s something wrong.
Does it sound familiar to you?
In an external hard drive that was stored carefully for the past fifteen years, I found films, series, photos, videos, and other paraphernalia. It had mp3 of the songs that I used to listen to; a quick glance and I saw Van Halen, Metallica, Megadeth, and Mötorhead among other artists. I opened a few of those files but most of them couldn’t be played or the bit rate was too low for me to even enjoy them. That’s what time does to you!
Truth be told, I was never an ardent follower of metal but the hard rock subgenre fit right into my alley. I had assumed that most of my Spotify’s playlist would be dominated by Rock but man! could I be any more wrong? I still listen to a lot of classics regularly and still nod my head a little. After checking the hard drive, I spent a few hours listening to the old music and later scoured the internet for a relevant dataset to see what’s happening in the school of rock; soon enough, I found one but it was for the metal bands not rock.
Oh well! it won’t be about headbanging but about violently shaking it.
Let’s go! 🤘
I found a dataset on Kaggle, you can access it here as well. It consists of information on metal bands from the inception of this subgenre of rock i.e. 1964. You can also use the one on death-metal from Metal-Archives which is scraped here.
After dropping the duplicates and those bands that didn’t have any country of origin, I moved forward to visualisations. Don’t worry! All those dropped ones weren’t popular as they had a really small number of fans(one of the variable).
I am not in a mood to use matplotlib, plotly, bokeh, or seaborn, so I will use Altair.
A lot of famous bands that I know are from the US and UK but northern Europe and Scandinavia aren’t far behind. Let’s investigate.
It’s quite easy to use Altair; I would recommend watching the first 40 minutes of Jake VanderPlas’(author) video.
You mention your dataframe as ‘source’ and use the attributes in it to build your graph. You can use options such as ‘sort’ within the field and can have the desired effect. No need to change the dataset exclusively.
As expected the US and UK have a lot of metal bands and the Scandinavian countries are at the top as well.
But is it the clear picture? USA, UK, and Germany have large populations, so maybe we need to find the number of metal bands per capita.
The chart was easy to formulate, I needed to create a temporary dataframe that would have the population of the countries as an attribute that can be used to conduct a more accurate analysis.
The Faroe Islands? Are they overdosing on Black Sabbath and Metallica? Well, they have 5 bands for a population of 50k, so for every 10k people, they have one band as we see in the chart. Also, Faroe Island is just next to Denmark and Sweden, so it is getting the inductive effect.
Had Vatican city have only a single band, it would have put all the other countries to shame(population of Vatican ~800 people).
What is the most popular subgenre in the metal world? I guess it will be heavy metal, thrash, black.
Please note that all the charts that we have done till now are bar charts, so the code above holds for all of them with minor changes.
To render these graphs in your notebook, VSCode, or Atom, please use:
I was right again!
Black, death, and heavy rule. Dark gods must be pleased.
How many metal bands are getting formed every year?
It seems the number of bands formed is decreasing over the years. Does that mean metal is dying?
Just because you’re not listening to it doesn’t mean it has met its maker. It would be fair to find bands forming and splitting every year to find the number of active bands per year.
Metal is a sub-genre of rock, it is harsher than hard rock which has found its audience in the general public and radio. You won’t hear metal songs on the radio, you have to find them on your own. If you won’t try, you’ll never find it and will assume that it is a dying culture.
What are the most popular metal bands in each country?
This is an interesting chart as the bar is labelled. You can choose any colour, you just need to specify the hex code and aren’t limited to the stylesheets of a particular library. You can formulate individual elements of the figure and can bind them like above by using (chart + text) to see the effects in a single chart.
I finally merged my dataframe with the data from Metal-Archives as it had the thematic information of the bands well. A simple word cloud tells a lot 🙂
Oh well! What were you expecting?
It’s a chart created using Folium and just another representation of the bar graph.
The verdict is clear.
Greenland, Iceland, Sweden, Norway, and Finland seem to be rocking pretty hard; Closer to the north pole, harder the head banging! 🙂
You can find the code to create the charts on my GitHub. As always, I can be reached on Twitter (@prashantmdgl9) or give a holler on LinkedIn. I would be interested to know what visualisations you would have created with the data.
Rock, hard rock, metal have gone through the litmus test of time; they aren’t well received by those who don’t consider them real music and even now there is constant comparison with old school, classical music which isn’t fair. But who really cares as long as you are enjoying it!
Also, I don’t understand the idea that why one has to restrict one’s music taste to one or the other form? I can enjoy a katcheri(an assembly of classical Carnatic musicians) as much as I like the music of Robert Plant.
So why choose?
After all, it’s all music and started with some guy banging two rocks together around 45,000 years ago.