<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Nathan Lang on Medium]]></title>
        <description><![CDATA[Stories by Nathan Lang on Medium]]></description>
        <link>https://medium.com/@nathanlang14?source=rss-40f83cdc98ed------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*mHPt-Tc4CHcvHUiKDCC4AQ@2x.jpeg</url>
            <title>Stories by Nathan Lang on Medium</title>
            <link>https://medium.com/@nathanlang14?source=rss-40f83cdc98ed------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 01:28:53 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@nathanlang14/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Hello “Covid”, World: Python Dash and R Shiny Comparison]]></title>
            <link>https://medium.com/data-science/hello-covid-world-python-dash-and-r-shiny-comparison-cc97afef9d82?source=rss-40f83cdc98ed------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc97afef9d82</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[r-shiny]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[dash]]></category>
            <category><![CDATA[r]]></category>
            <dc:creator><![CDATA[Nathan Lang]]></dc:creator>
            <pubDate>Wed, 10 Feb 2021 16:32:38 GMT</pubDate>
            <atom:updated>2021-02-10T16:42:19.024Z</atom:updated>
            <content:encoded><![CDATA[<h3>Hello, “Covid” World: Python Dash and R Shiny Comparison</h3><h4>Comparing Dash and Shiny in the most simple, but interactive dashboard implementation</h4><p>In the data science field, it is common for the final goal of a project to end in some sort of dashboard or web app to visualize your results. Both widely used data science languages, Python and R, each have their own packages to tackle this issue.</p><p>R has a single, main package named Shiny that makes it easy to build interactive web apps straight from R. On the other hand, Python has a few packages that are widely used depending on the intensity of the project.</p><p>The three main packages for Python web apps are Django, Flask, and Dash. Django is a high-level Python Web framework that encourages rapid development and clean, pragmatic design. Flask is a lightweight web application framework. It is designed to make getting started quick and easy, with the ability to scale up to complex applications.</p><p>Per the introduction of their website of Dash:</p><blockquote>Written on top of Flask, Plotly.js, and React.js, Dash is ideal for building data visualization apps with highly custom user interfaces in pure Python. It’s particularly suited for anyone who works with data in Python.</blockquote><blockquote>Through a couple of simple patterns, Dash abstracts away all of the technologies and protocols that are required to build an interactive web-based application. Dash is simple enough that you can bind a user interface around your Python code in an afternoon.</blockquote><p>Both Shiny and Dash have the common goal of making a web app dashboard with minimal to no web development skills like JavaScript. Although both do allow for better customization with the use of the JavaScript, HTML, CSS stack.</p><p>As many developers know, whenever you begin to learn a new programming language or tool the first thing you do is try to write a “Hello, World” program. This is to simply show the simplest form of the syntax and the inner workings of the new language.</p><p>This article takes that idea through the lens of a comparison between Python Dash and R Shiny to compare very simple dashboard web apps in both languages. Of course, it also is centered around COVID-19 data like the rest of the world has been for the past year.</p><p><strong>Disclaimer: I have experience with both Python and R, but I wanted to write this article while I am relatively unbiased and learning both Dash and Shiny for a fair comparison.</strong></p><h3>Barebones of an Interactive App for Dash and Shiny</h3><p>Dash and Shiny use a similar structure where you first define the overall interface and page layout in a hierarchical tree of components. This is then followed by functions that will change the layout depending on actions.</p><p>I chose to show the most basic interactive app because without interaction it is harder to depict the structure.</p><h4>Dash</h4><p>This is an example nearly line by line found <a href="https://dash.plotly.com/basic-callbacks">here</a>. It is important to note that Dash has amazing documentation found on that website.</p><p>In this example, we can see how the first step is to declare app.layout where the HTML divisions are defined in order from top to bottom of the web app. There is a dash core component(dcc) that takes in text and then the final html.Div is the output text. To link these two together, there is an @app.callback which is how Dash creates an action and reaction. The callback takes a component ID and property/value as input and passes it to the function, and then returns the result to the defined html.Div component.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/3d1d373aaf48c874f12b78c83021c0ff/href">https://medium.com/media/3d1d373aaf48c874f12b78c83021c0ff/href</a></iframe><h4>Shiny</h4><p>Here is a similarly easy, interactive Shiny app. Right away we notice a few similarities. First off, as mentioned previously, both programs inherit a similar structure where you define the UI and then the reactive elements. One difference is for this Shiny app all inputs are passed to the general function and each input is called upon by input$input_value.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/af216ec522e32ca7ad2c76061408e8bb/href">https://medium.com/media/af216ec522e32ca7ad2c76061408e8bb/href</a></iframe><p>As a quick comparison, we can see Shiny can deliver the same functionality with much less code, and less complex code. Part of the reason is Dash is built on Flask which is unopinionated meaning it is more open-ended and has no standard way for building because it is the developers’ decision. Furthermore, Dash has much more HTML integrated as seen in the above example.</p><p>Let’s further dive into this comparison by adding in some data to show common elements used such as a graph, dropdown menu, map, and slider.</p><h3>Element Comparison</h3><p>All of the code can be found on my GitHub <a href="https://github.com/Nathanlang14/HelloCovidWorld"><strong>here</strong></a>, but to conserve the length of this paper I am mainly going to focus on the specific code for the elements.</p><h4>Bar Graph and Dropdown Menu</h4><p>Dash comes with a Core Components Library, allowing easy declaration of common HTML components such as Dropdownsor Graphs. Since Plotly and Dash are very closely intertwined, almost all charting and graphing is done with Plotly’s graph_objects. So, once a dropdown value is selected, the dataset is then turned into a bar chart that overlays positive COVID-19 cases onto the total amount of tests to show how both have changed over time until the end of 2020.</p><p>One convenient thing about Plotly’s graph_objectsbeing connected with Dash is you automatically get Plotly features with no extra code. This includes zooms, hover feature, hover compare, download options, etc.</p><p>This code block depicts how to instantiate a dropdown menu and use the selected value to display a bar chart graph with Python and Dash.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/7286c186cb5d90c2d8765bae8e37828c/href">https://medium.com/media/7286c186cb5d90c2d8765bae8e37828c/href</a></iframe><p>This code produces the following bar chart which is determined by the dropdown menu value to depict either a single, specific state or aggregated total for all states.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IQIHQngAPGif_epujQX1_w.png" /><figcaption>Dash Dropdown Menu and Bar Chart Dashboard Example</figcaption></figure><p>Shiny also has a simple declaration for the UI elements, but when it comes to using the input you have to store it in a new variable to be able to compare to something else. This is because the input is a reactive element and can’t compare to a non-reactive variable, such as a string.</p><p>The nice thing about Shiny is most people who code in R already use the ggplot2package for all plotting so it is easy to implement the same plots right into the Shiny app.</p><p>Below is the R implementation of a similar program for dropdown menu and bar chart.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/3cb56d049f183f7e5a5161e1ce8f7130/href">https://medium.com/media/3cb56d049f183f7e5a5161e1ce8f7130/href</a></iframe><p>This code produces the Shiny version of the dropdown menu and bar chart to display total and positive cases over time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*szKlkFwfrVTXnPZY9Bh6Eg.png" /><figcaption>Shiny Dropdown Menu and Bar Chart Dashboard Example</figcaption></figure><p>Overall, the dropdown menu is nearly the same in terms of both looks and functionality. On the other hand, I have to give the right-out-of-the-box bar graph to Dash simply because it comes with plenty of attractive features from Plotly that are nice to have.</p><p>Aside: there is a <a href="https://plotly.com/r/">Plotly R library</a>, but it is not as naturally intertwined like Dash and Plotly are.</p><h4>Map and Slider</h4><p>In this section, we are going to create a slider of dates throughout 2020 from the date COVID-19 first appeared until December 31st. Then, once a date is picked, we will display the total case aggregated data onto a US map. Once again Plotly comes in clutch by having a choropleth map as part of their graph objects package.</p><p>The slider is another standard dccelement, but one huge complaint and inconvenience is it does not allow for dates as the input. So, we had to get a little creative by using the slider as just a number input starting at 0. Then, using the selected value along with the datetimePython package, we add that many days to the starting date. That being said, more time had to be spent to determine which marks represent the start of a new month.</p><p>This code block depicts how to instantiate a slider and display a map at the selected date with Dash.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/f1c3a92c152ccd3c4db77a735db4d664/href">https://medium.com/media/f1c3a92c152ccd3c4db77a735db4d664/href</a></iframe><p>This produces the following map that shows the total test cases by state in a choropleth US map by date selected with the slider.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4Zgb5TmiFdxb2PNNpQSfjg.png" /></figure><p>Once again Shiny has similar implementations, but I could not find any super-easy way to plot a US map. There were a few various examples where they use <a href="https://rstudio.github.io/leaflet/choropleths.html">Leaflet</a> but that seemed to have more customization than I wanted for a simple map. Therefore, I used a GitHub package from the <a href="https://github.com/UrbanInstitute/urbnmapr">Urban Institute called urbnmapr</a>.</p><p>I liked Shiny’s slider much more than Dash’s for this use case, specifically because it allowed for dates to be passed in as the value. So, there was no preprocessing needed to pass the selected value to the map. Furthermore, there was an animation option that made it possible for the map to progressively change by day.</p><p>This code block depicts how to instantiate a slider and display a map at the selected date with Shiny.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d2068f75d09a76ca65d7299cef87936a/href">https://medium.com/media/d2068f75d09a76ca65d7299cef87936a/href</a></iframe><p>This produces the following map that shows the total positive cases by state in a choropleth US map by date selected with the slider.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aIYkeOSKhI6z3iyeD21Ofw.png" /></figure><h3>Conclusion</h3><p>Overall, I hope you as a reader got a general gist of how simple components and reactive web apps are made for both Python Dash and R Shiny. This is my first official project for both so it was a learning process for me as well. For my initial thoughts, I see how both are useful in the data science field and I would say from my current point of view they are interchangeable enough to pick what you are comfortable with. That is if you are more comfortable with R stick with it, and similar for Python. I think it would be useful to learn both of course.</p><p>With that said, if I had to pick one to continue with, I would probably pick Dash for a few reasons. The first reason seen throughout this article, being intertwined with Plotly has a ton of advantages with features like zoom, pan, hover, etc. Secondly, Python is still a more popular language so for a project lifecycle it is probably easier to continue the dashboard stage in the same language. Lastly, Dash is built on top of Flask so if you ever wanted to extend your web app even further to a larger, more complex problem you would not have to necessarily restart.</p><p>Thank you all for reading! Feel free to let me know in the comments if anything surprised you or interested you. :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc97afef9d82" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/hello-covid-world-python-dash-and-r-shiny-comparison-cc97afef9d82">Hello “Covid”, World: Python Dash and R Shiny Comparison</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Analyzing Most Popular Spotify Artists Using Data Science]]></title>
            <link>https://medium.com/swlh/analyzing-most-popular-spotify-artists-using-data-science-814f26465370?source=rss-40f83cdc98ed------2</link>
            <guid isPermaLink="false">https://medium.com/p/814f26465370</guid>
            <category><![CDATA[spotify]]></category>
            <category><![CDATA[pandas]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[music]]></category>
            <dc:creator><![CDATA[Nathan Lang]]></dc:creator>
            <pubDate>Sat, 16 May 2020 01:05:42 GMT</pubDate>
            <atom:updated>2020-05-20T06:38:15.997Z</atom:updated>
            <content:encoded><![CDATA[<h4>Using Pandas to analyze Spotify Top 200 lists over the last 5 years</h4><h3>Overview</h3><p>Unfortunately, due to COVID-19, many of the activities that involve listening to music have been halted, such as exercising or commuting to work. No more window-down cruisin’ with the music blasting or attending music concerts/festivals. With that said, it doesn’t mean we can’t appreciate the success of the industry’s top artists. With my findings, I will be determining the most popular and most consistent artists, along with identifying the rising stars and let downs.</p><p>As anyone who has a passion for music and data science, I figured there is no better time to consider who has been at the top of the charts consistently over the last five years. I’m interested in exploring this question to determine what dictates popularity and consistency. In short, it could be total amount of listens or total amount of songs, but in this article I propose using the h-index to determine the top artists.</p><p>The h-index is usually used to express productivity and citation impact of scientists’ publications. For my question, I will be using the h-index to express the amount of songs an artist has with a <strong>h </strong>amount of one million listens or more. In simple terms, if an artist has an h-index of 10, that means they have 10 songs with at least 10 million listens or more.</p><p><strong>Side note</strong>: if you are not interested in the data science, but only the results feel free to skip ahead.</p><h3>Data Collection</h3><p>Spotify releases daily and aggregated weekly Top 200 charts at <a href="https://spotifycharts.com/regional">spotifycharts.com/regional</a>. Unfortunately, this only goes back to the start of 2017. This is where <a href="https://kworb.net/">kworb.net</a> comes in. I have spoken to Kworb on twitter, and they have been scraping the data since August 10th, 2014. So, the dataset I used is the aggregated U.S. Daily Top 200 charts from that date through May 14, 2020. It is important to note that these totals do not include the time spent off the chart.</p><p>Pandas provides a nice, easy way to read HTML tables.</p><pre>import pandas as pd<br>df = pd.read_html(&#39;<a href="https://kworb.net/spotify/country/us_daily_totals.html&#39;)[0">https://kworb.net/spotify/country/us_daily_totals.html&#39;)[0</a>]</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*rjq7IAIwSfPmUQ7uMZhC_A.png" /><figcaption>Head of Dataframe</figcaption></figure><h3>Data Cleaning and Processing</h3><p>As any data scientist will tell you, data cleaning is the most important (yet tedious) process of any project, regardless of project size. Luckily, pandas interpreted the dataframe rather well, limiting the amount of cleaning that was needed.</p><p>While there are many interesting questions that can be explored with this dataset, the only parts I am interested in are the artist, amount of songs that were on the Top 200 list during this timeframe, and total amount of streams while on the Top 200 list.</p><p>I now create a new pandas dataframe <em>artist_stats</em> that will only display artist, song title, and streams. Since I will be using the streams in millions, I divide and round the total streams by one million.</p><pre>clean_df = pd.DataFrame()<br>clean_df[[&#39;Artist&#39;,&#39;Title&#39;]] = df[&#39;Artist and Title&#39;].str.split(&#39; - &#39;,expand = True, n = 1)<br>clean_df[&#39;Streams&#39;] = (df[&#39;Total&#39;]/1000000).round(decimals = 0).astype(int)</pre><p>With simple pandas manipulations, I am able to get a clean dataset consisting of only the data I want.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*I9B8BeLkLwUJH_fzn7EoTg.png" /><figcaption>Head of Cleaned Dataframe</figcaption></figure><p>Given this new dataframe, I am now able to calculate the h-index by using a groupby transform and a lambda function. This single line of code is very powerful. First, it groups by artists and then it calculates how many songs have more streams than their rank by streams.</p><pre>clean_df[&#39;h-index&#39;] = clean_df.groupby(&#39;Artist&#39;)[&#39;Streams&#39;].transform(lambda x: ( x &gt;= x.rank(ascending=False, method=&#39;first&#39;)).sum())</pre><p>Finally, I will use pandas’ awesome groupby-aggregate function to get the final dataframe that I will analyze. I am then able to obtain total streams, song count, the h-index, and the h-index average for each artist. The h-index average is the percentage of songs that are attributed to the h-index for an artist.</p><pre>artist_stats = clean_df.groupby(&#39;Artist&#39;).agg({&#39;Streams&#39;:&#39;sum&#39;, &#39;Title&#39;:&#39;nunique&#39;, &#39;h-index&#39;:&#39;mean&#39;}) </pre><pre>artist_stats.rename(columns = {&#39;Title&#39;:&#39;Song_Count&#39;}, inplace = <strong>True</strong>)</pre><pre>artist_stats[&#39;h-index_avg&#39;] = (100*artist_stats[&#39;h-index&#39;] / artist_stats[&#39;Song_Count&#39;]).round(decimals = 2)</pre><pre>artist_stats.sort_values(&#39;h-index&#39;, ascending = <strong>False</strong>, inplace = <strong>True</strong>)</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/511/1*tpX2y1MxVJbDg_dslyoiCw.png" /><figcaption>Top 15 Sorted By the h-index</figcaption></figure><p>All of the code I used for this project can be viewed on my github <a href="https://github.com/Nathanlang14/Spotify_Artist_Analysis">here</a>.</p><h3>The Fun Part (Results)</h3><p>Now that I have a clean dataset, let’s do some analyzing. Before diving in, I want to clarify that this dataset only consists of songs and their streams while on the <strong>Spotify’s Top 200 Daily U.S. chart</strong>. Also, this list does not take into account when people are featured on a song. Only the artist who released the song is accredited the statistics.</p><h4>Most Popular</h4><p>It is clear that <strong>Drake </strong>and <strong>Post Malone</strong> are huge front runners in the popularity contest. There is no surprise with Drake being up top as he entered this timeframe already very popular. Drake also debuted 152 songs in the last five years on Top 200, while Post Malone only had a third at 50 songs. I am actually more impressed with Post Malone because his first album <em>Stoney</em> was only released in December 2016. So, that means all the other artists, including Drake, had over two years to produce popular music. Talk about execution! Post Malone did top the charts with the most popular song “Rockstar” and the third most popular song with “Sunflower”. “Rockstar” was #1 on the charts for a whopping 124 days straight, the longest of any song during this time!</p><p>Drake and Post Malone both had over 6 billion stream listens, while the next highest was Ariana Grande at 2.4 billion which is only 40% of those two. Drake is clearly the most consistent artist of our time with his first studio album, <em>Thank, Me Later,</em> was released in 2010, four years prior to this dataset. Drake has undoubtedly been dominant over the last ten years. With that said, I am going to give most popular award to Post Malone.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bsIKq7makcLL9wmLeegAbQ.jpeg" /><figcaption>Post Malone and Drake: <a href="https://urbanislandz.com/2019/12/12/post-malones-hollywoods-bleeding-ranks-biggest-album-of-2019/">source</a></figcaption></figure><h4>Rising Stars</h4><p>The most impressive rising star for me is <strong>Billie Eilish. </strong>While she is only 18 years old, she has produced plenty of popular songs. She has an h-index of 20 with only 24 songs that broke into the Top 200. For those who may have glanced over the h-index meaning, that means Billie has 20 songs with at least 20 million listens. Now this is nothing less than amazing as that means only four of her songs that made it to the Top 200 have not blown up. Billie’s most popular song is “bad guy”.</p><p>Billie’s second most popular song is “lovely” which actually features my next rising star: <strong>Khalid</strong>. Similar to Billie, Khalid is young at 22 years old and has featured 36 songs on the Top 200, with an h-index of 17. Khalid is currently known and recognized by his two most popular songs “Location” and “Young, Dumb &amp; Broke”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tG5XzUcOb3PKqN-rmrujVA.jpeg" /><figcaption>Billie Eilish and Khalid: <a href="https://www.youtube.com/watch?v=wjj2upnfBI0">source</a></figcaption></figure><p>My last rising star maybe a little controversial to fit into this category: <strong>Travis Scott</strong>. This is a tough call putting him in this category due to his popularity and presence already, but I am putting him here strictly because I think he has the potential to be the next Drake. Yes, that is a bold statement. But, he has only had 32 songs on the Top 200 in last five years making him the second lowest behind Billie. Out of these 32, 53% have over 17 million streams with “SICKO MODE” having 475 million and “goosebumps” having 472 million. Only Drake and Post Malone have two songs with over 450 million streams while on the Top 200. Although it is important to note that Drake is featured on “SICKO MODE”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lIt-Nk8SsgZMsvchZ4JdSg.jpeg" /><figcaption>Travis Scott: <a href="https://urbanislandz.com/2018/08/05/travis-scott-addresses-trans-model-astroworld-cover-controversy/">source</a></figcaption></figure><h4>Let Downs</h4><p>There are only two people with over 75 songs that appeared on the Top 200 beside Drake. These two people are also the only people with an h-index of over 10 with an h-index average of under 15%. They are <strong>Future </strong>and <strong>Logic</strong>.</p><p>Future has debuted 136 songs that made it onto this list, while only 16 of them had over 16 million streams. Future is also a let down because, out of his top five streamed songs, 4 out of 5 had a feature on it and 2 of them were Drake.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tmvIfngE2jJwLM7uf9eyxg.jpeg" /><figcaption>Future: <a href="https://urbanislandz.com/2019/01/18/listen-future-drops-new-album-the-wizrd/">source</a></figcaption></figure><p>Similarly, Logic had 87 songs on this list and only 11 had 11 million streams or more. Logic only had one song with over 75 million streams and it was “1–800–273–8255” which had 320 million. He is as close to a popular one-hit-wonder as anyone.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*2cLTvruOEsEngnzwM6xKsQ.jpeg" /><figcaption>Logic: <a href="https://www.revolt.tv/2019/4/4/20825426/logic-makes-history-as-the-first-rapper-with-a-no-1-best-selling-novel">source</a></figcaption></figure><h4>Honorable Mentions (R.I.P.)</h4><p><strong>Juice WRLD</strong> and <strong>XXXTENTACION </strong>were two of my personal favorites. Both had bright futures in the music industry and unfortunately passed away at young ages. Both Juice WRLD (21) and XXX (20) were rising stars that took the music industry by storm with their new genre emo-rap. Juice WRLD’s “Lucid Dreams” had 547 million streams and he had 7 other songs with over 100 million streams. Similarly, XXX had “SAD!” with 518 million streams and 6 other songs with over 100 millions streams (one at 99 million). It is without a doubt, that both Juice WRLD and XXXTENTACION would have had continued success in the industry.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pJb4Q0bZrOp5o0zwnd-ntg.jpeg" /><figcaption>XXXTENTACION and Juice WRLD: <a href="https://www.youtube.com/watch?v=m5FG5PK9qXA">source</a></figcaption></figure><p>Thank you all for reading! Feel free to let me know in the comments if anything surprised you or if there are any other questions about this dataset that interest you. :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=814f26465370" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/analyzing-most-popular-spotify-artists-using-data-science-814f26465370">Analyzing Most Popular Spotify Artists Using Data Science</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>