<?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 Quantum Reports on Medium]]></title>
        <description><![CDATA[Stories by Quantum Reports on Medium]]></description>
        <link>https://medium.com/@quantumreports?source=rss-6b46e676c1a4------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Ayqed1nA3sX4R6yXxo9lNQ.jpeg</url>
            <title>Stories by Quantum Reports on Medium</title>
            <link>https://medium.com/@quantumreports?source=rss-6b46e676c1a4------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 00:08:12 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@quantumreports/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[My Last Umoja Hack Africa Experience as an Undergrad: The Good, the Bad, and the Ugly.]]></title>
            <link>https://quantumreports.medium.com/my-last-umoja-hack-africa-experience-as-an-undergrad-the-good-the-bad-and-the-ugly-dbfbaabfb6a4?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/dbfbaabfb6a4</guid>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Thu, 25 Jan 2024 09:10:44 GMT</pubDate>
            <atom:updated>2025-06-17T16:35:34.537Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/481/1*5vru5gsaI6Le09rePLrpbw.jpeg" /></figure><p>Umoja Hack Africa is the largest inter-university hackathon in Africa, held annually over two days. It brings together talented AI enthusiasts from universities across the continent, including beginners, intermediates, and advanced levels. The hackathon features three levels of challenges: beginner, intermediate, and advanced, with the latter usually focusing on a deep learning problem.<br>The hackathon is held virtually on the <a href="https://zindi.africa/">Zindi</a> platform, and a leaderboard is employed to rank participants as they proffer solutions to the problem statement as solo competitors or as part of a team.</p><p>This year was my first time participating in Umoja Hack Africa and I competed in the <a href="https://zindi.africa/competitions/umojahack-africa-2023-intermediate-challenge">intermediate challenge</a>. The intermediate challenge was a classification-based problem. This challenge provided insights into crypto-jacking and its effect on computing speed and computational costs. The objective of the challenge is to classify network activity from multiple websites into either crypto jacking or non-crypto jacking by utilizing features associated with both network-based and host-based information. I wasn’t a solo participant as I competed as part of a team named- <a href="https://zindi.africa/competitions/umojahack-africa-2023-intermediate-challenge/leaderboard/teams/marines">Marines</a>.</p><p>My teammates at this year’s Umoja Hack Africa were <a href="https://zindi.africa/users/Edifon">Edifon Emmanuel Jimmy</a> and <a href="https://zindi.africa/users/ImonEmmanuel/">Emmanuel Imonmion</a> and together, we called ourselves the Marines. We represented our University (University of Benin) at the hackathon. We were a team with diverse strengths in exploratory data analysis, features engineering, and advanced model development. During the hackathon, I focused on performing domain-based features engineering, leveraging extensive research on the problem statement to gain insights into the target and independent variables.</p><h4>My Team’s Approach to the Hackathon</h4><p>As a rule of thumb, when participating in an ML competition, it’s crucial to develop a baseline model with minimal data preprocessing. This approach provides a starting point for evaluating the data’s potential performance. Once the baseline model is built, it can be refined by processing the data effectively, thereby enhancing the accuracy of the model.<br>Using the XGBoost algorithm, our team submitted our initial model, with an F1 Score as our error metric, resulting in leaderboard scores ranging from 0 to 1. Unfortunately, our first submission scored only 0.256666666, far behind the other participants. We then tried the CatBoost algorithm, which resulted in an even lower score. Upon investigation, we discovered an unusual issue: the ID column in our submission file was different from the ID column in the test file. Correcting the issue, we submitted again and achieved a remarkable score of 0.940594059, placing us in the top 20 on the public leaderboard.<br>Throughout the hackathon, we continuously enhanced our model by performing exploratory data analysis, feature engineering, and feature importance analysis. My team implemented advanced machine learning techniques such as cross-validation, hyperparameter tuning, stacking, and ensemble modelling to further improve our public leaderboard score. On the final day of the hackathon, we leveraged ensembling techniques to combine our best approaches. Our efforts resulted in an impressive final leaderboard score of 0.966887417 and a private leaderboard score of 0.968107725.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bZxN72VewcVwn9xaYhq2Nw.png" /></figure><p>Reflecting on my experiences at UHA ’23, I can say that they were diverse and impactful. They have enriched my problem-solving skills and influenced my perspective on competitive data science. To start on a positive note, let me share the good experiences.</p><h4>The Good</h4><p>Despite encountering several obstacles during the hackathon, my team managed to secure the 47th position out of 328 active participants/teams and 437 enrolled participants/teams. In addition to enhancing my machine learning skills, I also learned the importance of teamwork in achieving goals, particularly in a competitive setting such as UHA. Working in a data science team, whether in research or product, was another valuable experience. My participation in UHA inspired me to delve deeper into research papers and notebooks, improve my Kaggle presence, and optimize my data science workflow.</p><h4>The Bad</h4><p>Participating in Umoja Hack Africa 2023 was not without its challenges. From day one, my team faced constant electricity outages, forcing us to relocate to our university’s public park to continue hacking. At one point, we had to leave due to weather conditions and welfare concerns. Due to the lack of reliable electricity, we were only able to make a few submissions on the first day. This hindered our ability to implement our ideas early and perform extensive experimentation with the data.</p><h4>The Ugly</h4><p>Now, the ugly……hmmmmm…<br>It’s crucial to remember the importance of cross-validation when building models to avoid overfitting. During the hackathon, we improved our leaderboard score by building stack models on the second day. Unfortunately, we didn’t have the time to implement a cross-validation strategy, causing us to hack blindly. As a result, we missed out on a good submission that utilized cross-validation and ended up choosing some submissions without it. This decision cost us the 22nd position on the leaderboard. <br>A note of advice: <strong><em>Always trust your cross-validation score.</em></strong></p><p>One of our shortcomings during the hackathon was neglecting to conduct a thorough EDA between the train and test sets. This oversight deprived us of valuable insights that could have led to substantial improvements in our leaderboard performance.</p><p>Although we were unable to execute all the ideas we had due to time constraints, the weekend of hacking was still an incredible experience.</p><h4><strong>Reflections</strong></h4><p>Representing my university on a continental platform was a privilege that provided me with a wealth of experience. Throughout the hackathon, I built new relationships, made friends, overcame personal fears, pushed through obstacles, and gained valuable knowledge. As a message to all students intending to dive into the field of data science and artificial intelligence, I encourage you to take advantage of the massive opportunities available in this field. It requires grit, consistency, and patience, but the rewards are worth it.</p><p>As a participant in Umoja Hack Africa- Zindi, I observed that there is a wealth of AI talent in Nigeria with budding ideas, despite limited resources. However, I believe that more technical and knowledge investments are needed to support Nigeria’s ML ecosystem. I suggest that the organizers of Umoja Hack Africa- Zindi focus on making such investments to support and nurture Nigeria’s AI talent and their innovative ideas. By doing so, we can help accelerate the growth of Nigeria’s ML ecosystem and support the country’s rise as a leader in AI and data science in Africa.</p><p>If you’re interested in seeing a sample submission for the hackathon, I’ve uploaded a notebook to Kaggle. You can find the link below. If you find it helpful and insightful, kindly upvote it!</p><p><a href="https://www.kaggle.com/danielakhabue/uha-23-submission-notebook">UHA &#39;23 Submission Notebook</a></p><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio: <a href="https://brave-sea-0769e151e.4.azurestaticapps.net/"><strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue">Follow me on Twitter</a> <em>if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dbfbaabfb6a4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why Data Science?]]></title>
            <link>https://quantumreports.medium.com/why-data-science-6241a0289509?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/6241a0289509</guid>
            <category><![CDATA[why-data-science]]></category>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Thu, 25 Jan 2024 09:06:35 GMT</pubDate>
            <atom:updated>2025-06-17T16:37:20.253Z</atom:updated>
            <content:encoded><![CDATA[<p>Have you ever been fascinated by patterns in data, or intrigued by the untold stories in data? Or even had this savior complex when you see data going through a rough patch?<br>If you have a yes to any of my questions, then you just might want to give the field of data science a try.<br>What really is Data Science?</p><p><strong>Data Science</strong>:<br>Data science is the domain of study that deals with vast volumes of data using modern tools and techniques to find unseen patterns, derive meaningful information, and make business decisions. Data science uses complex machine learning algorithms to build predictive models.</p><p><a href="https://www.simplilearn.com/tutorials/data-science-tutorial/what-is-data-science">What is Data Science? A Simple Explanation and More</a></p><p><strong>The Journey so far as a Data Scientist<br></strong>Entering the Data Science ecosystem as a college student was a tedious one considering the not-so-clear approach to the field. I had to peruse a lot of online materials (video and text) and attend several online boot camps and online courses with little or no success.<br>My way out of the quagmire was deliberately reaching out to mentors and colleagues. With the help of these individuals, I have gotten experience in Visualized Data Analysis using tools like Microsoft Excel, Microsoft PowerBI, and Google Data Studio for my dashboard analysis. <br><em>P.S. I would share my experience with these tools in subsequent posts</em>.<br>Data Science is so broad a field that it covers such skill sets as data analysis, data engineering, machine learning, and statistics.</p><p>I just built a machine learning model for predicting <strong>house prices</strong> using <strong>regression analysis</strong>. The data set was gotten from <a href="https://zindi.africa/competitions/free-ai-classes-in-every-city-hackathon-2023">zindi</a>. <br>Skillsets and Tools used:<br>1. Data cleaning<br>2. Data Analysis<br>3. Programming(Python)<br>4. Data Visualization<br>5. Data Wrangling<br>6. Kaggle Notebook</p><p>Let me share with you some of the processes I underwent.<br><strong>Data Cleaning</strong>: This stage is the most tedious of all processes involved in building a model because the cleaner and better your data, the more accurate your model gets. Employing my programming and data visualization skills, I cleaned the dataset by removing duplicates and <a href="https://www.scribbr.com/statistics/outliers/">outliers</a> and checking for empty cells.</p><p><strong>Data Wrangling</strong>: I wrangled the data by converting several categorical data into binary formats using one-hot encoding for easy data processing.</p><p><strong>Data Analysis and Visualization</strong>: I then analyzed the data for relationships between the different columns for patterns using visualization plots like bar charts, histoplots, boxplots, distribution plots, and Scatterplots.</p><p><strong>Modeling</strong>: After analyzing my data, I made the necessary decisions concerning my data set. I then used a linear regression model imported from sci-kit learn to build my model. The accuracy of the model was tested with a test data set and documented in the notebook. <br>This project and its notebook can be gotten from my <a href="https://brave-sea-0769e151e.4.azurestaticapps.net/">portfolio</a>.</p><p>Back to why data science, I guess you know the answer already.</p><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio: <a href="https://brave-sea-0769e151e.4.azurestaticapps.net/"><strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue">Follow me on Twitter</a> <em>if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6241a0289509" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Driving Electoral Integrity with Data Visualization: How a Power BI Dashboard Can Revolutionize…]]></title>
            <link>https://quantumreports.medium.com/driving-electoral-integrity-with-data-visualization-how-a-power-bi-dashboard-can-revolutionize-d76aa8f66239?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/d76aa8f66239</guid>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Thu, 25 Jan 2024 08:58:37 GMT</pubDate>
            <atom:updated>2025-06-17T16:36:25.813Z</atom:updated>
            <content:encoded><![CDATA[<h3>Driving Electoral Integrity with Data Visualization: How a Power BI Dashboard Can Revolutionize Your Election Process.</h3><p>The principles of accountability and transparency are crucial to ensuring the integrity and legitimacy of any electoral process. As a manager of a team of tech experts, I had the privilege of leading the development of an e-voting solution aimed at addressing the key challenges faced by traditional voting systems. My team was able to overcome the issues of overvoting, ease of access, effectiveness, and time constraints by leveraging the power of technology to create a more efficient and streamlined voting process.</p><p>Transparency is essential to building trust and confidence in the electoral process, and our e-voting solution is needed to deliver on this front as well.</p><h4>Analytics wins it all</h4><p>To address this issue, we took advantage of the data analytics workflow and came up with an innovative approach to enhance the transparency of the e-voting results. We integrated data visualization that ensures the integrity and transparency of the results. This groundbreaking solution significantly improved the transparency and accountability of the electoral process and set a new standard for e-voting systems.<br>I chose Microsoft Power BI for creating the dashboard since it is versatile, easy to use, and has a myriad of options for data visualization.</p><p>In summary, our e-voting system had not only solved the traditional challenges of voting but also demonstrated the power of technology to revolutionize the electoral process. Our innovative approach had set a new benchmark for transparency and accountability in the electoral process and has opened up new possibilities for future advancements in this critical area.</p><p>To utilize Microsoft Power BI for visualizing electoral data, you need to get acquainted with Power BI Desktop. The tutorial below gives a simple and easy way to use the application and create compelling visuals.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FPXqFXG4rzXE%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DPXqFXG4rzXE&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FPXqFXG4rzXE%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/f7c6347baf6ae111103e78717ebd1de4/href">https://medium.com/media/f7c6347baf6ae111103e78717ebd1de4/href</a></iframe><h4>Why a Dashboard?</h4><blockquote><a href="https://support.google.com/analytics/answer/1068216?hl=en#zippy=%2Cin-this-article">Dashboards are a collection of widgets that give you an overview of the reports and metrics you care about most. Dashboards let you monitor many metrics at once, so you can quickly check the health of your accounts or see correlations between different reports. Dashboards are easy to create, customize and share.</a></blockquote><p>Dashboards help you see figures in more descriptive forms, and also bring a collection of reports together to provide useful insights.</p><h4>Steps to Implement an Electoral Dashboard</h4><h4>1. Getting the Data:</h4><p>The e-voting software utilized MongoDB to store the election candidates’ data and their corresponding vote tallies for each position. Once the election concluded, the data for each position was efficiently retrieved as a CSV file, allowing for seamless analysis and review.</p><h4>2. Importing the Data and Building the Dashboard:</h4><p>Just like the video above, the data was imported into the Power BI desktop. Key information like the number of Registered voters, the number of Voters, and the number of Non-voters, were gotten and indicated on the dashboard. Appropriate visuals were used to describe the winners of the various executive positions.</p><h4><strong>3.</strong> Publishing the Dashboard:</h4><p>The dashboard was made available to the public by exporting it as a PDF file. A snapshot of the dashboard is presented below for reference:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1n-JuBUKgHpZddmy7C9dRg.png" /></figure><h4><strong>Conclusion</strong></h4><p>Transparency was a key objective in presenting the election results, and the implementation of a dashboard achieved this goal by showcasing figures obtained from the database. The dashboard was a resounding success, effectively displaying the election outcomes in a user-friendly, straightforward, and innovative manner. I highly recommend trying it out!</p><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio: <a href="https://brave-sea-0769e151e.4.azurestaticapps.net/"><strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue">Follow me on Twitter </a><em>if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d76aa8f66239" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploratory Data Analyses: NPFL’s Teams and Standings.]]></title>
            <link>https://quantumreports.medium.com/exploratory-data-analyses-npfls-teams-and-standings-fa0a513f02dd?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/fa0a513f02dd</guid>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Mon, 21 Nov 2022 03:50:02 GMT</pubDate>
            <atom:updated>2025-06-17T16:34:03.595Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/255/1*yJpaSOS_dDx9ZZp07-gBOg.png" /></figure><p>The NPFL (Nigeria Professional Football League) is Nigeria’s top-tier league and a very unpopular one amongst Nigerians. The League is managed by the LMC (League Management Company) and consists of the best 20 domestic teams in Nigeria where each team plays a total of 38 matches in a season except on special occasions as was seen in this project.</p><p>I decided to embark on this after I had seen several analyses on European leagues especially the English premier league (EPL). While viewing these analyses, I began to ask questions about the reasons for the absence of dashboards, notebooks, and models around the NPFL (I mean, it’s a very interesting league with a lot of news, upsets, and shockers). In my research, I found out one of the reasons was the lack of an accessible data repository of the league’s stats.</p><p>I then decided to show the world what happens in Nigeria’s top-tier domestic league using interactive dashboard analysis. Following this decision, I embarked on the most gruesome step in the project, Data Acquisition/Cleaning. For the sake of other Data Scientists and analysts who would want to use the dataset, I decided to mine league data across several seasons.</p><p>Like other leagues’ analyses, I felt it would be nice to see stats like:</p><ol><li>Team positions</li><li>Average goals scored per match</li><li>Number of away goals per team in a season</li><li>Number of home goals per team in a season</li><li>Number of draws by a team in a season</li></ol><p>With these in mind, I followed the following steps for my project.</p><ul><li>Data Acquisition/Data Validation/Data Cleaning</li><li>Data Transformation</li><li>Data Modeling</li><li>Data Visualization (Power BI Dashboard)</li><li>Dashboard Publishing</li></ul><p><strong>Data Acquisition/Data Validation/Data Cleaning</strong></p><p>At this stage of the project, I gathered NPFL stats from various online blogs and betting sites using several web scraping techniques and tools like BeautifulSoup and Requests libraries. You can view a replica of my code <a href="https://github.com/Apprentice-doa/Data_Scraping_with_Python">here</a>. This stage was the most difficult as I had to deal with a lot of wrong data, outdated data, and inconsistent data formats. Also, I found out that I couldn&#39;t find some interesting stats like:</p><ol><li>Conversion rates of shots</li><li>Number of completed passes</li><li>Distance covered by players in a match.</li></ol><p>After pulling a lot of data for the different seasons from various sources, I compared them across different news and information sites like Wikipedia to validate them.</p><p>It was at this stage I found out the following about the various seasons:</p><p><strong>2018/2019 NPFL Season</strong></p><p>Unlike the normal, this season had 24 teams compete in the league due to the league being suspended in the previous season. So, no team was relegated while 4 teams were promoted to the league from the NNL (Nigeria’s second-tier football league). The 24 teams were divided into 2 groups. The top three teams from each of the groups are to qualify for playoffs where a champion is to emerge.</p><p>Below are the results of the groups and playoffs:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gbvpPsy8N2vQCeFoyZWaCg.png" /><figcaption>NPFL 2018/2019 Group A results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_FlFAE7w-VxBSFNxcSOIjA.png" /><figcaption>NPFL 2018/2019 Group B results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lui51juInyGCWsMI9zsv-Q.png" /><figcaption>NPFL 2018/2019 Championship Playoffs results.</figcaption></figure><p>During my analysis, I observed that Delta Force bought their league’s spot from Kada City Fc after the latter had played only 11 games in the season (Source: <a href="https://en.wikipedia.org/wiki/2019_Nigeria_Professional_Football_League">Wikipedia</a>). I had to effect this change in my dataset because a lot of data sources did not capture this transaction.</p><p><strong>2019/2020 NPFL Season</strong></p><p>The 2019/2020 season was suspended due to the covid-19 pandemic which meant that the total number of games played by a team was less than 38. Also, it was noticed that the total number of games played by the teams was not uniform hence, the points system was not used for ranking rather, the points per total game played(PPG) system was used to rank the teams on the league standings. In a situation where teams had the same PPG, the team with the less number of games played was ranked higher. If the clubs played the same number of games, the club with the higher goal difference is then ranked higher.</p><p>Below is the final league table for the season:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZLMLhQiR0Gzis5JfGJyEkA.png" /><figcaption>NPFL 2019/2020 results.</figcaption></figure><p>During my analysis, I observed that Kwara United bought their league’s spot from Delta Force Fc after the latter had played only 7 matches in the season (Source: <a href="https://en.wikipedia.org/wiki/2019%E2%80%9320_Nigeria_Professional_Football_League">Wikipedia</a>). I had to effect this change in my dataset because a lot of data sources did not capture this transaction.</p><p><strong>NPFL 2020/2021 Season</strong></p><p>During the compilation of this season’s dataset, I observed inconsistent data across different data sources, especially in league standings. Further investigation revealed that in a matchday 34 fixture between Jigawa Golden Stars and Rivers United which initially ended in a 2–1 loss to Rivers United, Jigawa Golden Stars fielded an ineligible player. The issue wasn’t resolved immediately so when the fixture was awarded in favour of Rivers United with a 3–0 win, a lot of online data sources did not capture the change (Source: <a href="https://en.wikipedia.org/wiki/2020%E2%80%9321_Nigeria_Professional_Football_League">Wikipedia</a>).</p><p>Below is the final league table for the season:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CO4I_lwwljIkEX579-cf6g.png" /><figcaption>NPFL 2020/2021 results.</figcaption></figure><p><strong>NPFL 2021/2022 Season</strong></p><p>This season was made up of 20 teams with 38 matches played by each team.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0bjJnnCsZ925dCV-0P2CGA.png" /><figcaption>NPFL 2021/2022 results.</figcaption></figure><p><strong>Data Transformation</strong></p><p>After cleaning the data, I then conducted feature engineering on the different data sets were created new features namely: W, L, D, GF/Goal Scored, GA, Season, and Ordinal Position.</p><p>W: This means the number of games a team won at the season ends. It was gotten by adding both HW and AW (Home win and Away win) together.</p><p>L: This means the number of games lost by a team at the end of the season. It was gotten by adding HL and AL (Home Loss and Away Loss) together.</p><p>D: This means the number of games drawn by a team at the end of the season. It was gotten by adding HD and AD (Home Draw and Away Draw) together.</p><p>GF/Goal Scored: This means the number of goals scored by a team at the end of the season. It was gotten by adding HF and AF (Home For and Away For) together.</p><p>GA: This means the number of goals conceded by a team at the end of the season. It was gotten by adding HA and AA (Home against and Away against) together.</p><p>Season and Ordinal Position features were added for analytical explanations on the dashboard.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JXTIyIY6BOQMjPipkbvk1A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*skNdT78n6kDtgY4DXkyquw.png" /><figcaption>Code to create the ordinal positions</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_74XvczPRLTALNm4pIRW5Q.png" /><figcaption>The final column after custom is merged with Pos-Copy</figcaption></figure><p><strong>Calculated Measures</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z4QMJuu5z3qMDvGQ_OqU_Q.png" /><figcaption>Dax expression to calculate the total number of matches played in the season.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wGZK3XfLBxp5-RNyhc09iQ.png" /><figcaption>Dax expression to calculate the Average goals scored per match in the season.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*91MiCe9etbplGsBUOYnswQ.png" /><figcaption>Dax expression to calculate the total number of goals scored in the season.</figcaption></figure><p>The following are the tables after the transformation of the dataset:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_b4T6WZGaTm_YojakHmQfw.png" /><figcaption>NPFL 2018/2019 Group A results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IA64ZhVR23Qz3C5-trIeEQ.png" /><figcaption>NPFL 2018/2019 Group B results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jM5xU20G7Y2zCixhP_yDaQ.png" /><figcaption>NPFL 2018/2019 Championship Playoffs results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*54RjMGLpzaTTeiQsUiKzHA.png" /><figcaption>NPFL 2019/2020 results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*swlD2rQStMIHYq10vP7N1w.png" /><figcaption>NPFL 2020/2021 results.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*CDnw0XZpcdNxJU1U6WUepg.png" /><figcaption>NPFL 2021/2022 results.</figcaption></figure><p><strong>Data Modeling</strong></p><p>With my dataset in the right format, I went ahead to find unique relationships between my datasets by creating a new table called “Teams”. This table is to relate all my season tables together.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2LvYeOGqIqdwanRq8v0Q6Q.png" /></figure><p><strong>Data Visualization (Power BI Dashboard)</strong></p><p>Check out my notebook EDA on the seasons:</p><p><a href="https://www.kaggle.com/danielakhabue/code">https://www.kaggle.com/danielakhabue/code</a></p><p><strong>Some Insights (2018/2019 Season)</strong></p><ul><li>Enyimba FC and Kano Pillars qualified for the CAF Champions League (Africa’s version of the UEFA Champions League) while Enugu Rangers qualified for CAF Confederations Cup as the third-placed team.</li><li>Niger Tornadoes became Nigeria’s second representative in CAF Confederations Cup after winning the Aiteo Cup (FA Cup).</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WdyysnQ9D0p8Blgex3Bqgg.png" /></figure><p><strong>Some Insights (2019/2020 Season)</strong></p><ul><li>246 matches were played in the league with 494 goals scored. Plateau United, Enyimba, Rivers United, Kano Pillars, Nasarawa United, and Sunshine Stars were the top teams with the most goals in the Season.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qiNBOwnTpgMNI_F7rFcdcQ.png" /></figure><p><strong>Some Insights (2020/2021 Season)</strong></p><ul><li>380 matches were played in the season while 791 goals were scored. Akwa United and Rivers United came out as Nigeria’s representatives at the CAF Champions League.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zt4WMMCVlUuwNortq6OuWg.png" /></figure><p><strong>Some Insights (2020/2021 Season)</strong></p><ul><li>380 matches were played with 779 goals scored in the season. Remo stars qualified for CAF Confederations Cup as the third-placed team.</li><li>Due to CAF’s deadline for submission of the country’s representatives at the continental level by NFF, no Aiteo cup was organised. Therefore, the fourth-placed team, Kwara United, was chosen as Nigeria’s second representative at the CAF Confederations Cup.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kVAkjJMxPP4PFrptKqweKQ.png" /></figure><p><strong>Dashboard Publishing</strong></p><p>The dashboard was published using Power BI Service for interactiveness.</p><p>For more insights, check out the interactive dashboard:</p><p><a href="https://bit.ly/npfl_dashboard">Power BI Report</a></p><p><strong>Recommendations:</strong></p><ol><li>Due to the amount of time it took to get the league’s data for analysis, I recommend that a public repository be created specially by the LMC (League Management Company) and the NFF (Nigeria Football Federation) to store important league data to checkmate other sources of data and avoid misinformation.</li><li>I recommend that more investment should be made in streaming NPFL matches which will ramp up revenues for the clubs and prevent the sale of their league positions due to financial reasons. This act affects the spirit of the game.</li><li>Massive investments should be made in sports analytics in the NPFL. It is a good way of bringing the attention of the public to the league and investors, as a side effect. Hardware like semiautomatic cameras, Local Positioning Systems (LPS), Global Positioning Systems (GPS) (source: <a href="https://barcainnovationhub.com/monitoring-technologies-for-sports-analysis/">Barca innovation hub</a>) and software analytical tools like Microsoft Power BI should be invested in for real-time analysis of football matches. Nigeria has a rich repository of Sports analytics talents so, we do not need to go out of the country to source for talent. This is another way of creating jobs in the league. NPFL Clubs should also invest in analytics technologies to brand their club’s identities and promote their clubs.</li></ol><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio:<a href="https://shared-wallet-2ed.notion.site/Daniel-Akhabue-2589d68da9134e88a60d907df97e7938"> <strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue/">Follow me on Twitte<em>r</em></a><em> if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fa0a513f02dd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Becoming a Contributor to the Open-Source Data Community.]]></title>
            <link>https://quantumreports.medium.com/becoming-a-contributor-to-the-open-source-data-community-aaf7257adda7?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/aaf7257adda7</guid>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Mon, 21 Nov 2022 03:37:34 GMT</pubDate>
            <atom:updated>2025-06-17T16:32:57.889Z</atom:updated>
            <content:encoded><![CDATA[<h4>Steps to starting as an open-source contributor in the data space.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_8bnBKyxhnR4ARqyFmNzgQ.jpeg" /></figure><p>An open-source community is a community of contributors globally who develop, test and modify open-source projects. Anyone who wishes to volunteer to contribute to an open-source project is allowed to do so. This is what makes it open-source. An open-source community is formed when a person or group of persons developing software releases the source code of the software to the public. They also give others permission to copy, modify, and redistribute the source code under certain terms. These terms are stated in the license of the open-source software.</p><p><a href="https://opensource.com/resources/what-open-source">What is open source?</a></p><p>It is important to point out that participating in an open-source project is a good reference for your resume. Having experience in contributing to an open-source community can serve as leverage before recruiters.</p><p>Let’s look at some of the steps to take when starting an open-source contribution.</p><ol><li>Contributing to Data Repositories. I have discovered that one of the easy ways of contributing to the open-source community is by contributing datasets to public data repositories. As a data professional, easy access to quality data is of great importance which is why public repositories like Kaggle are available for uploads and downloads of datasets.</li></ol><p><a href="https://www.kaggle.com/">Kaggle: The World&#39;s AI Proving Ground</a></p><p>2. Get used to using version control software like Git for publishing your data projects. Understanding how version control works will help you know how to contribute to open-source projects. This is because a lot of open-source projects are hosted on GitHub.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FRGOj5yH7evk%3Fstart%3D1562%26feature%3Doembed%26start%3D1562&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DRGOj5yH7evk&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FRGOj5yH7evk%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/7eb5f138a95b6a4a625a0c6cb1663aea/href">https://medium.com/media/7eb5f138a95b6a4a625a0c6cb1663aea/href</a></iframe><p>3. Another tip for participating in an open-source data project is to read the documentation of the project. A funny approach is to check for spelling or grammatical errors in the README file of the project and correct them.</p><p>4. Learn how to write codes in OOP (Object-Oriented Programming). This is important because open-source projects are written with code re-usability in mind and, OOP covers that. Another advantage of learning how to write codes in OOP is that it helps in the modularity of code. This means that you can focus on a code issue without tampering with other parts of the code.</p><p>5. Finally, check the issues in an open-source repository for code issues you can solve. Doing this will keep you updated on what is happening in the project.</p><p>Here is a list of open-source projects you can try your hands with as a data professional.</p><ol><li>Pandas- <a href="https://github.com/pandas-dev/pandas">https://github.com/pandas-dev/pandas</a></li><li>Scikit Learn- <a href="https://github.com/scikit-learn/scikit-learn">https://github.com/scikit-learn/scikit-learn</a></li><li>Tensorflow- <a href="https://github.com/tensorflow/tensorflow">https://github.com/tensorflow/tensorflow</a></li><li>Numpy- <a href="https://github.com/numpy/numpy">https://github.com/numpy/numpy</a></li><li>Pytorch- <a href="https://github.com/pytorch/pytorch">https://github.com/pytorch/pytorch</a></li></ol><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio:<a href="https://shared-wallet-2ed.notion.site/Daniel-Akhabue-2589d68da9134e88a60d907df97e7938"> <strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue">Follow me on Twitte<em>r </em></a><em>if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aaf7257adda7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Is hacking together your first python package difficult?]]></title>
            <link>https://quantumreports.medium.com/is-writing-your-first-python-package-difficult-2577a61f97ee?source=rss-6b46e676c1a4------2</link>
            <guid isPermaLink="false">https://medium.com/p/2577a61f97ee</guid>
            <dc:creator><![CDATA[Quantum Reports]]></dc:creator>
            <pubDate>Mon, 21 Nov 2022 03:36:30 GMT</pubDate>
            <atom:updated>2025-06-17T16:36:57.015Z</atom:updated>
            <content:encoded><![CDATA[<h3>Is Hacking Together Your First Python Package Difficult?</h3><h4>Tips for writing an excellent python package.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/381/1*ncHA6wiq1tJ39XM-Tvzfng.jpeg" /></figure><p>Python language is a robust programming language that is broadly used in different technology niches. It is a language that finds immense relevance in web development, data analytics, data science, machine learning, and artificial intelligence.</p><p>It is widely sought after for its plethora of packages and libraries which allow for efficient coding.</p><p>There are a lot of things to know when writing a python package. Don’t fret, they are easy to grasp as long as you show interest. Let us look at the definition of some terms used around python packages.</p><ol><li><strong>Module:</strong> A module is a file that contains functions, classes, or variables. A module always has a <strong><em>.py</em></strong> extension.</li><li><strong>Package: </strong>A package is a folder containing related python modules and an <strong><em>__init__.py </em></strong><em>(</em>pronounced<em>- double underscore eeneet double underscore eeneet dot pie).</em> The __init__.py file is required so that the python interpreter can recognize that it is a package. It is a unit of distribution for sharing your code with users and developers. The __init__.py file can be empty.</li><li><strong>Library:</strong> A library is a collection of packages or/and modules for reusability within a community. A library has to be installed before it is imported for its reusable code.</li></ol><p>There are important things to note when writing a python package.</p><p><strong>First, </strong>understand the concepts surrounding <em>object-oriented</em> <em>programming</em> like <em>abstraction</em>,<em> inheritance, encapsulation, and polymorphism.</em></p><p><a href="https://betterprogramming.pub/python-for-beginners-object-oriented-programming-3b231bb3dd49">Python for Beginners — Object-Oriented Programming</a></p><p><strong>Second,</strong> ensure that each function you create serves a particular purpose i.e. it is solving a specific problem.</p><p><strong>Third,</strong> ensure that your function names are descriptive enough for the task it is executing. Also, make certain that your function names are distinct from each other.</p><p><strong>Fourth</strong>, write meaningful and concise documentation using in-line comments, doc-strings comments, or read-me documentation.</p><p><a href="https://medium.com/swlh/guide-to-documenting-your-python-code-4683d33bdf73">Guide to Documenting Your Python Code</a></p><p><strong>Fifth,</strong> Modularize your codes for easy implementation and fixing of bugs. This means you separate your codes into different modules.</p><p><strong>Finally,</strong> ensure that you create a new and empty python file in your folder of modules and name it<strong> <em>__init__.py</em></strong><em>.</em></p><p>To see an example of creating a python package, view the video content below.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F4uosDOKn5LI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D4uosDOKn5LI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F4uosDOKn5LI%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/3a02a581790ecc572afc129a989c50f4/href">https://medium.com/media/3a02a581790ecc572afc129a989c50f4/href</a></iframe><p><strong><em>Before you leave:</em></strong></p><p>If you liked this article, don’t forget to give me a few claps, <a href="https://quantumreports.medium.com/">follow me</a> and thus receive all updates about new publications.</p><p><strong>About the author:<br><em>Daniel Akhabue</em></strong>: Data Scientist, writer, <a href="https://www.kaggle.com/danielakhabue">Kaggle Expert</a>, and Team Lead at <a href="https://twitter.com/DSnaiUniben?t=UJyUuv7ros3YfH69q9W6A&amp;s=08"><strong>DSN Uniben</strong></a></p><p><strong><em>Email: </em></strong>akdaniel0009@gmail.com</p><p>Check out my portfolio:<a href="https://shared-wallet-2ed.notion.site/Daniel-Akhabue-2589d68da9134e88a60d907df97e7938"> <strong><em>Daniel Akhabue</em></strong></a></p><p><a href="https://x.com/_danielakhabue">Follow me on Twitte<em>r</em></a><em> if you want even more content.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2577a61f97ee" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>