<?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 Modupe Giwa on Medium]]></title>
        <description><![CDATA[Stories by Modupe Giwa on Medium]]></description>
        <link>https://medium.com/@modupe-giwa?source=rss-189c00762943------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*RCD_ifJFfGIIGQsqvzuStQ@2x.jpeg</url>
            <title>Stories by Modupe Giwa on Medium</title>
            <link>https://medium.com/@modupe-giwa?source=rss-189c00762943------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 03:19:16 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@modupe-giwa/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[I Thought I Knew What I Wanted to Do With My Career]]></title>
            <link>https://modupe-giwa.medium.com/i-thought-i-knew-what-i-wanted-to-do-with-my-career-198748a019ea?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/198748a019ea</guid>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[short-story]]></category>
            <category><![CDATA[storytelling]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Wed, 07 Oct 2026 10:16:17 GMT</pubDate>
            <atom:updated>2026-10-07T10:16:17.752Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OApswwfRzJ739ypgvrPfdw.png" /></figure><p>When I was in secondary school, I knew one thing for certain: I wanted to be a career woman. I didn’t necessarily know what career I wanted, but I knew what I wanted my life to look like. I imagined myself working in a corporate office, wearing heels, carrying a laptop, walking into meetings and being taken seriously. I wanted to be ambitious. I wanted to be good at what I did. I wanted to become the kind of woman who had built something meaningful for herself.</p><p>I think I’ve always had that picture of myself. For a while, I thought that meant I was going to become a banker. We had tenants who were bankers, and I remember being fascinated by them. They always looked so put together. The suits, the shoes, the way they carried themselves. Everything about them seemed so professional and impressive to me.</p><p>At some point, I even told my late grandmother that I wanted to become a banker. Then secondary school happened. I ended up in the science class mostly because my friends were going there. At that point, I didn’t have a carefully thought-out career plan. I was just following the path in front of me. Somewhere around SS1 and SS2, I realised that I actually loved Mathematics. So naturally, I started thinking that maybe I should study Mathematics at university. But somehow, I ended up studying Industrial and Production Engineering.</p><p>Looking back, the choice actually made sense for the person I was becoming. I didn’t want something purely technical. I was interested in how things worked, but I was equally interested in the management side of things. Industrial and Production Engineering felt like a middle ground — a combination of engineering, operations, management and problem-solving. So I chose it. And for a while, I thought I had finally figured it out.</p><p>During university, I became interested in engineering management and data analytics. I started imagining myself doing a master’s in engineering management and building my career around that. It felt like a plan. Then I graduated. And, as it turns out, having a plan doesn’t necessarily mean knowing what you’re doing.</p><p>I had different opportunities after graduation, and eventually chose a role that was heavily rooted in engineering. I thought, Okay. Maybe this is it. But within the first few months, I started noticing something. I didn’t particularly enjoy the technical, hands-on parts of the work.</p><p>What I found myself enjoying were the parts that involved analysing data, conducting risk assessments, working with documentation, looking for patterns, understanding what was going wrong and trying to figure out why. I somehow became the person people could come to when something needed to be analysed, documented or organised.</p><p>And I kept thinking: Wait. Maybe this is what I actually enjoy. It wasn’t engineering itself that excited me as much as I initially thought it would. It was the problem-solving. The analysis. The business side of things. The opportunity to sit with information, understand what it was saying and turn it into something useful.</p><p>That realisation changed the way I started thinking about my career. I began learning data analytics seriously. I took courses in Excel, Power BI and SQL. I worked on projects. I started paying more attention to the kinds of problems I naturally enjoyed solving.</p><p>Then I started exploring Data Analysis, revenue operations and go-to-market roles. And now, somehow, I’m here. I recently moved into a trade marketing role, and I’m curious about where it will take me. Not because I suddenly have everything figured out. Actually, quite the opposite. Because I’m beginning to understand that my career might not be one straight line.</p><p>Sometimes I still think about consulting. Sometimes I think about operations. Sometimes I think about analytics. Sometimes I think about strategy. And sometimes I wonder whether I should eventually go back to the things I originally thought I wanted.</p><p>There are so many possible directions that it can be difficult not to feel like you are making the wrong choice every time you choose one path over another. And I think that is the part of career conversations we don’t talk about enough.</p><p>The uncertainty. The changing of your mind. The realisation that something you thought you would love isn’t actually for you. The excitement of discovering something new, only to wonder whether you’re going to change your mind about that too.</p><p>I used to think that knowing what I wanted to do meant choosing a career and sticking to it. Now I’m not so sure. Maybe the girl who told my grandmother she wanted to be a banker wasn’t wrong. Maybe the girl who chose engineering wasn’t wrong either. Maybe the woman learning data analytics wasn’t starting over.</p><p>Maybe none of those versions of me was wrong. They were just making decisions with the information they had at the time. And maybe that is why I don’t think I can confidently say what I want to do with my career anymore.</p><p>I still know that I want to build something. I still know that I am ambitious. I still see myself walking into that office, putting on those heels, sitting at that table and becoming the kind of woman I used to admire from a distance. I just don’t know exactly what my job title will say when I get there.</p><p>And, for the first time, I’m beginning to think that maybe that’s okay.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=198748a019ea" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Supply Chain Analytics Dashboard with Python and Power BI]]></title>
            <link>https://modupe-giwa.medium.com/building-a-supply-chain-analytics-dashboard-with-python-and-power-bi-0bd3ac4fbd7d?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/0bd3ac4fbd7d</guid>
            <category><![CDATA[supply-chain]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[power-bi]]></category>
            <category><![CDATA[dax]]></category>
            <category><![CDATA[dashboard]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Mon, 10 Aug 2026 20:19:31 GMT</pubDate>
            <atom:updated>2026-08-10T20:19:31.541Z</atom:updated>
            <content:encoded><![CDATA[<p>Supply chain data can tell you much more than how many orders were placed. It can reveal where deliveries are falling behind, which markets drive revenue, which products generate the most profit, and where operational performance needs attention.</p><p>For this project, I worked with the <a href="https://www.kaggle.com/datasets/shashwatwork/dataco-smart-supply-chain-for-big-data-analysis">DataCo Global Supply Chain dataset</a>, containing over 180,000 transaction records, to build an end-to-end supply chain analytics dashboard.</p><p>My goal was simple: turn raw supply chain data into insights that can support better business and operational decisions.</p><h3>The Process</h3><p>I started by using Python and Pandas to clean and prepare the dataset. This involved removing unnecessary columns, standardising the data, transforming date fields, and creating additional variables required for the analysis.</p><p>The cleaned dataset was then imported into Power BI, where I built an interactive dashboard across four key areas:</p><ol><li>Executive Overview: overall sales, profit, orders, customers, and delivery performance.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/973/1*c6J2F0DpXv4Ylw25JYCJHA.png" /><figcaption>Executive Overview Page</figcaption></figure><p>2. Order Management: order status, processing time, cancellations, suspected fraud, and order trends.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/978/1*SRyJFFzzCaBb8sfgdcxsmg.png" /><figcaption>Order Management Page</figcaption></figure><p>3. Product &amp; Customer Analysis: product profitability, customer segments, average order value, repeat customers, and geographic distribution.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/976/1*9VUBaAD9FmVqHXupoBZ_Mw.png" /><figcaption>Product &amp; Customer Page</figcaption></figure><p>4. Logistics: shipment volumes, shipping modes, delivery times, and on-time versus late deliveries.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/656/1*AP_rPFIjgAC-Tq7mX0CWlg.png" /><figcaption>Logistics Page</figcaption></figure><h3>What Did the Data Reveal?</h3><ol><li>The business generated $36.8 million in sales and approximately $4 million in profit, with a 10.8% profit margin. It also processed approximately 65,800 orders from 20,700 customers.</li><li>Europe emerged as the largest market by sales, generating approximately $10.9 million, followed closely by LATAM at $10.3 million.</li><li>Product performance also showed some clear leaders. Fishing was the most profitable product category, followed by Cleats and Camping &amp; Hiking.</li><li>One of the most important findings, however, came from the logistics analysis.</li></ol><p>Only 42.7% of shipments were delivered on time, while 57.3% experienced late delivery.</p><p>This suggests that while the business is generating significant sales and order volume, delivery reliability remains an important operational challenge.</p><h3>Recommendations</h3><p>Based on the analysis, I would recommend that the business:</p><ol><li>Investigate the causes of the high late-delivery rate and identify the markets, routes, and shipping modes contributing most to delays.</li><li>Review shipping-mode allocation and determine whether orders can be assigned more effectively based on required delivery times.</li><li>Prioritise high-performing product categories while monitoring demand and availability to prevent fulfilment issues.</li><li>Develop market-specific logistics strategies, particularly for high-volume markets where delivery improvements could affect a large number of orders.</li></ol><h3>Tools Used</h3><ol><li>Python</li><li>Pandas</li><li>Jupyter Notebook (Google Colab)</li><li>Power BI</li><li>DAX</li></ol><h3>Final Thoughts</h3><p>What I enjoyed most about this project was moving beyond simply creating charts.</p><p>The dashboard helped connect financial performance, customer behaviour, order management, product performance, and logistics operations in one place.</p><p><strong>Check out the GitHub repository for the dataset, Python notebook, Power BI dashboard, and project files </strong><a href="https://github.com/Modupegiwa/DataCo-Supply-Chain-Analytics"><strong>here</strong></a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0bd3ac4fbd7d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[MTN Nigeria Q1 2025 Customer Retention Dashboard Using Python and Power BI]]></title>
            <link>https://modupe-giwa.medium.com/mtn-nigeria-q1-2025-customer-retention-dashboard-using-python-and-power-bi-a723b126a2f9?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/a723b126a2f9</guid>
            <category><![CDATA[exploratory-data-analysis]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[pandas]]></category>
            <category><![CDATA[power-bi]]></category>
            <category><![CDATA[numpy]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Tue, 21 Jul 2026 16:25:24 GMT</pubDate>
            <atom:updated>2026-07-21T16:25:24.550Z</atom:updated>
            <content:encoded><![CDATA[<p>In the telecommunications industry, customer retention is one of the biggest drivers of growth. Acquiring new customers is important but understanding why existing customers leave and identifying opportunities to improve customer experience can have an even greater impact on long-term revenue.</p><p>For telecom companies operating in highly competitive markets like Nigeria, factors such as network experience, pricing, customer satisfaction, and product offerings can influence whether customers remain loyal or switch to competitors.</p><p>In this project, I analyzed a simulated MTN Nigeria Customer Churn &amp; Usage Dataset (Q1 2025) to understand customer behavior, identify churn drivers, and build an interactive business intelligence dashboard using Python and Power BI.</p><h4>Objectives</h4><p>The goal was to answer key business questions:</p><ol><li>Who are the customers most likely to churn?</li><li>What factors contribute to customer attrition?</li><li>Which products and customer segments generate the most revenue</li><li>What retention strategies can improve customer loyalty?</li></ol><h4>Dataset Overview</h4><p>The dataset was sourced from Kaggle and contains customer-level information including:<br> • Customer demographics<br> • Location<br> • Device type<br> • Subscription plans<br> • Data usage behavior<br> • Revenue information<br> • Customer satisfaction ratings<br> • Churn status and reasons</p><p>The dataset contains 974 customer records with multiple transactions/devices per customer, allowing analysis of both customer behavior and product performance.</p><h4>Tools &amp; Technologies Used</h4><ol><li><strong>Python (Google Colab): </strong><br> • Data cleaning<br> • Data quality checks<br> • Feature engineering<br> • Exploratory data analysis</li></ol><p>Key libraries used: Pandas, NumPy and Matplotlib</p><p><strong>2. Google BigQuery</strong></p><p>Used as a cloud data warehouse to store and query the transformed dataset and establish a connection with Power BI for reporting.</p><p>3. <strong>Power BI</strong></p><p>Power BI was used to create the final interactive dashboard.</p><p>Key techniques applied:<br> • Data modelling<br> • DAX calculations<br> • KPI development<br> • Time-based comparisons<br> • Interactive slicers<br> • Business storytelling</p><h4>Data Preparation &amp; Feature Engineering</h4><p>Before building the dashboard, I performed several data preparation steps:</p><ol><li><strong>Data Cleaning</strong><br> • Checked for missing values<br> • Removed duplicate records<br> • Standardized categorical fields<br> • Converted date fields into usable formats<br> • Validated revenue calculations</li></ol><p>2.<strong> Feature Engineering</strong></p><p>Additional analytical field was created:</p><ul><li><strong>Age Groups</strong></li></ul><p>Customers were segmented into:<br> • Gen Z<br> • Millennials<br> • Gen X<br> • Seniors</p><p>This helped identify demographic trends in customer behavior.</p><h4><strong>Dashboard Design</strong></h4><ol><li><strong>Executive KPI Overview</strong></li></ol><p>At the top of the dashboard, I included 7 KPI cards that provide an instant snapshot of business performance:<br> • Total Revenue: ₦199M<br> • Total Customers: 496<br> • Total Active Customers: 350<br> • Churn Rate: 29.4%<br> • Retention Rate: 70.6%<br> • Average Customer Tenure: 31 months<br> • Average Data Usage: 99 GB</p><p>To make the dashboard more actionable, each KPI includes a Month-over-Month comparison, allowing users to quickly identify whether performance is improving or declining.</p><p><strong>For example:</strong><br> • Customer growth increased by 35.9% compared to the previous month.<br> • Churn rate increased by 1.4%, signalling a negative trend that requires attention.<br> • Retention rate declined by 0.6%, reinforcing the need for customer retention initiatives.</p><p><strong>2. Interactive Filtering</strong></p><p>To encourage self-service analysis, I added interactive slicers that allow users to explore the data without modifying the report.</p><p>Users can filter the dashboard by:</p><p>• Month</p><p>• State</p><p>• Gender</p><p>• MTN Device</p><p>I also included a Reset Filters button so users can instantly return the report to its default view after exploring different scenarios.</p><p><strong>3. Dynamic Trend Analysis</strong></p><p>One of my favourite features is the dynamic KPI trend chart.</p><p>Instead of creating multiple line charts, I used Power BI Field Parameters so users can switch between different business metrics using a single visual.</p><p><strong>The available metrics include:</strong></p><ul><li>Total Revenue</li><li>Total Cost</li><li>Profit Margin</li><li>Churn Rate</li><li>Active Customers</li></ul><p>This keeps the dashboard clean while allowing executives to analyse multiple performance indicators over time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1005/1*ajBpi2KU_SYGEMAh2p89Sw.png" /><figcaption>The Power BI Dashboard</figcaption></figure><p><strong>4. Customer Retention Analysis</strong></p><p>To better understand customer behaviour, I included several visuals that highlight churn patterns across different dimensions.</p><ul><li>A donut chart shows the proportion of active and churned customers, providing an immediate view of customer retention.</li><li>A stacked bar chart breaks down churn by Age Group and Gender, making it easier to identify demographic segments with higher attrition.</li><li>Another horizontal bar chart highlights the states with the lowest customer churn, helping identify regions where customer retention strategies appear to be most effective.</li><li>Finally, I analysed customer behaviour across MTN devices and subscription plans, enabling comparisons between Mobile SIM users, Broadband MiFi users, and 5G Broadband Router customers to identify products associated with higher or lower churn.</li></ul><h4>Key Insights</h4><p>1.<strong> Customer Churn Remains a Major Retention Challenge</strong></p><p>The analysis identified a significant percentage of customers leaving the network.</p><p>Customers with lower satisfaction scores showed a higher likelihood of churn, highlighting the relationship between customer experience and retention.</p><p>2. <strong>Customer Tenure Provides a Retention Opportunity</strong></p><p>Customer tenure analysis showed that retention strategies should not only focus on new customers but also customers approaching important lifecycle stages.</p><p>3. <strong>Product Usage Can Identify Growth Opportunities</strong></p><p>Customers with high data consumption represent valuable opportunities for upselling premium plans.</p><h4>Recommendations</h4><ol><li>Create proactive retention campaigns targeting customers with declining satisfaction scores before they leave.</li><li>Introduce loyalty rewards and personalized offers for customers reaching key tenure milestones.</li><li>Use customer usage behavior to recommend personalized data bundles and broadband products.</li></ol><h4>Challenges &amp; Lessons Learned</h4><p>One major lesson from this project was understanding that good analytics requires more than creating visuals.</p><p>Metrics such as churn rate and retention rate require careful consideration of:<br> • Data structure<br> • Customer-level aggregation<br> • Appropriate DAX calculations</p><p>For example, because customers could have multiple devices, customer metrics required using DISTINCTCOUNT(Customer ID) rather than counting transaction rows.</p><h4>Designing for Decision-Making</h4><p>A dashboard should not only answer “what happened?” but also help answer:<br> • Why did it happen?<br> • Who is affected?<br> • What action should be taken?</p><p>This influenced the dashboard design by focusing on actionable insights rather than displaying every available metric.</p><h4>Conclusion</h4><p>This project improvedmy understanding of how data analytics can support customer retention strategies.</p><p>Through Python-based analysis and Power BI visualization, I transformed raw customer data into a business intelligence solution that highlights churn patterns, customer behavior, and revenue opportunities.</p><p>The biggest takeaway is that analytics is not just about creating dashboards — it is about turning data into decisions.</p><h4>Project Links</h4><p>🔗 <a href="https://www.kaggle.com/datasets/oluwademiladeadeniyi/mtn-nigeria-customer-churn">Kaggle Dataset</a></p><p>🔗 <a href="https://github.com/Modupegiwa/MTN-Customer-Churn-Analysis-Q1-2025">GitHub Repository</a></p><h4>Tools:</h4><p>Python | Pandas | NumPy | Google BigQuery | Power BI | DAX | Power Query</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a723b126a2f9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How I Built an Automated Daily News Briefing System with n8n and Airtable]]></title>
            <link>https://modupe-giwa.medium.com/how-i-built-an-automated-daily-news-briefing-system-with-n8n-and-airtable-f7b4633803ce?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/f7b4633803ce</guid>
            <category><![CDATA[gmail-api]]></category>
            <category><![CDATA[n8n]]></category>
            <category><![CDATA[airtable]]></category>
            <category><![CDATA[workflow-automation]]></category>
            <category><![CDATA[workflow]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Mon, 20 Jul 2026 18:01:13 GMT</pubDate>
            <atom:updated>2026-07-20T19:20:55.166Z</atom:updated>
            <content:encoded><![CDATA[<p>Keeping up with technology news can be overwhelming, so I built an automated system that delivers a curated tech news briefing directly to subscribers’ inboxes every morning at 8:00 AM.</p><h4>Objectives</h4><ol><li>Collect subscribers through an online form.</li><li>Store subscriber information in a database.</li><li>Automatically fetch the latest technology news every morning.</li><li>Personalise and send the newsletter to every subscriber.</li><li>Keep track of when each subscriber last received an email.</li></ol><h4>Tools Used</h4><p>The entire workflow is powered by n8n, Airtable, and the Gmail API, making it a lightweight, low-cost automation project.</p><p><strong>If you’d like to receive the daily briefing yourself, you can subscribe here:</strong></p><p>🔗 Sign up <a href="https://mogiwa.app.n8n.cloud/form/9a78d1d3-cb99-4c17-92db-70ae25124821">here</a></p><h4>The Workflow Architecture</h4><p>The automation consists of two independent workflows running on the same n8n canvas.</p><pre>PATH 1 – Subscriber Signup<br><br>[Form Submission]<br>        │<br>        ▼<br>[Create Record in Airtable]</pre><pre>PATH 2 – Daily Newsletter (Runs Every Day at 8:00 AM)</pre><pre>[Schedule Trigger]<br>        │<br>        ▼<br>[Search Subscribers]<br>        │<br>        ▼<br>[Read RSS Feed]<br>        │<br>        ▼<br>[Code (Format Newsletter)]<br>        │<br>        ▼<br>[Send Email (Gmail)]<br>        │<br>        ▼<br>[Update Airtable Record]</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dHFiFgk75d2GyboVJN6jPA.png" /><figcaption>The Workflow Architecture</figcaption></figure><p>Although these workflows live on the same canvas, they operate independently. One handles new subscriptions, while the other runs automatically every morning to send the newsletter.</p><h4>How It Works Step-by-Step</h4><ol><li><strong>Capturing New Subscribers</strong></li></ol><p>I created a simple n8n-hosted form where visitors enter their name and email address.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/973/1*RDFYAr_mhCQDQ39qODAAKA.png" /><figcaption>The n8n form</figcaption></figure><p>When the form is submitted:</p><ul><li>The subscriber’s details are captured instantly.</li><li>A new record is created in Airtable.</li><li>The subscriber is now included in future daily newsletter runs.</li></ul><p>Using Airtable makes it easy to manage subscribers without setting up a traditional database.</p><p><strong>2. Sending the Daily Newsletter</strong></p><p>Every day at <strong>8:00 AM</strong>, the second workflow starts automatically.</p><p><strong>A. Schedule Trigger: </strong>The Schedule Trigger launches the workflow at the same time every morning.</p><p><strong>B. Retrieve Subscribers</strong></p><p>The workflow searches Airtable for all active subscribers.</p><p>Each subscriber record includes:</p><ul><li>Name</li><li>Email Address</li><li>Last Email Sent</li></ul><p><strong>C. Fetch the Latest Tech News</strong></p><p>The RSS Read node retrieves the latest articles from a technology RSS feed (such as TechCrunch).</p><p>This removes the need to manually search for news each day.</p><p><strong>D. Build the Email</strong></p><p>A JavaScript Code node:</p><ul><li>selects the latest articles,</li><li>formats them into clean HTML,</li><li>personalises the greeting using each subscriber’s name, and</li><li>prepares the email for delivery.</li></ul><p><strong>E. Send Personalised Emails</strong></p><p>Using the Gmail API, n8n sends the newsletter individually to every subscriber.</p><p>Each recipient receives a clean, personalised email containing the latest technology headlines.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*f6YTn1bpQexZebZ9VD7xjA.png" /><figcaption>The Personalized Email</figcaption></figure><p><strong>F. Update Airtable</strong></p><p>After an email is successfully sent, Airtable is updated with the current date in the <strong>Last Email Sent</strong> field.</p><p>This provides a simple delivery history for every subscriber.</p><h4>Challenges I Solved</h4><p>Building the workflow wasn’t entirely straightforward. Here are a few issues I encountered and how I resolved them.</p><ol><li><strong>Keeping Independent Workflows Together</strong></li></ol><p>The subscriber signup process and the daily newsletter run at completely different times.</p><p>Instead of creating separate workflows, I kept both on the same n8n canvas while allowing each to have its own trigger.</p><p>This keeps everything organised without introducing unnecessary complexity.</p><p><strong>2. Preserving Item Relationships (</strong><strong>pairedItem)</strong></p><p>Inside the JavaScript Code node, n8n lost track of which subscriber each email belonged to.</p><p>By including the pairedItem property in the output, I preserved the relationship between each subscriber and their generated email.</p><p>This prevented downstream execution errors.</p><p><strong>3. Updating the Correct Airtable Record</strong></p><p>Initially, Airtable updates failed because the workflow wasn’t referencing the correct record.</p><p>The solution was to dynamically use:</p><pre>{{ $json.recordId }}</pre><p>instead of hardcoding record IDs.</p><p>This ensures every subscriber’s record is updated correctly after their email is sent.</p><h3>Final Thoughts</h3><p>This setup automates both audience capture and content delivery without ongoing manual effort.</p><p>Without any manual intervention, the workflow:</p><ul><li>captures new subscribers,</li><li>stores them in Airtable,</li><li>fetches the latest technology news,</li><li>generates a personalised HTML newsletter,</li><li>sends it through Gmail, and</li><li>records the delivery date for each subscriber.</li></ul><p>The same architecture can easily be adapted for:</p><ul><li>company newsletters,</li><li>blog updates,</li><li>AI news digests,</li><li>market reports,</li><li>industry-specific briefings, or</li><li>internal team updates.</li></ul><p>It was a great exercise in workflow automation and a practical demonstration of how n8n can orchestrate multiple services into a reliable end-to-end solution.</p><p><strong>You can find the complete n8n workflow JSON and documentation in my </strong><a href="https://github.com/Modupegiwa/Daily-News-Briefing"><strong>GitHub repository</strong></a><strong>.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f7b4633803ce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Building a Revenue Operations Dashboard with BigQuery and Google Data Studio (Looker)]]></title>
            <link>https://modupe-giwa.medium.com/building-a-revenue-operations-dashboard-with-bigquery-and-google-data-studio-looker-6ff3f3e9b508?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/6ff3f3e9b508</guid>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[bigquery]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[looker]]></category>
            <category><![CDATA[google-data-studio]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Wed, 08 Jul 2026 22:25:44 GMT</pubDate>
            <atom:updated>2026-07-08T22:25:44.656Z</atom:updated>
            <content:encoded><![CDATA[<p>Revenue Operations (RevOps) helps organisations align marketing, sales, and customer success by using data to improve decision-making.</p><p>To build my analytics portfolio and gain hands-on experience with RevOps reporting, I built an interactive dashboard using Google’s public TheLook E-commerce dataset.</p><h4>About the Dataset</h4><p>TheLook E-commerce is a public dataset available in Google BigQuery. It simulates the operations of an online fashion retailer and contains several years of transactional and behavioral data.</p><p>The dataset includes information across seven tables:</p><ul><li>Users (customer profiles and demographics)</li><li>Orders</li><li>Order Items</li><li>Products</li><li>Inventory Items</li><li>Website Events</li><li>Distribution Centers</li></ul><p>These tables capture customer profiles, website interactions, purchasing behavior, product information, inventory, and fulfillment activities, making the dataset suitable for end-to-end business analysis.</p><h4>Objectives</h4><p>The project was designed to answer the following business questions:</p><ul><li>How is the business performing financially?</li><li>Which marketing channels generate the most conversions?</li><li>Which products contribute the most to revenue and profit?</li><li>How valuable are the customers, and how does customer growth change over time?</li></ul><h4>Tools Used</h4><ul><li><strong>Google BigQuery: </strong>Data cleaning, transformation, and SQL view creation.</li><li><strong>SQL: </strong>Data preparation and business metric calculations.</li><li><strong>Google Data Studio (Looker): </strong>Interactive dashboard development and data visualization.</li></ul><h4>Data Preparation</h4><p>To simplify reporting and improve dashboard performance, I created 3 analytical SQL views in BigQuery instead of connecting every dashboard directly to the raw tables.</p><p>The project includes dedicated views for:</p><ul><li>Executive Revenue</li><li>Marketing Operations</li><li>Customer &amp; Growth</li></ul><p>Creating purpose-built views reduced unnecessary joins in Looker Studio and made the dashboards easier to maintain.</p><h4>Dashboard Overview</h4><h4>1. Executive Revenue Dashboard</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*gUzNwcsfGaQLDcHLQ9mZ2A@2x.jpeg" /></figure><p>The Executive Revenue dashboard provides a high-level view of business performance.</p><p>It includes KPIs such as:</p><ul><li>Total Revenue</li><li>Total Profit</li><li>Total Orders</li><li>Total Customers</li><li>Average Order Value</li><li>Profit Margin</li></ul><p>The dashboard also includes monthly revenue and profit trends, revenue by product category, revenue by department, geographic revenue distribution, and a table highlighting top-performing products.</p><p>This view allows stakeholders to quickly understand overall business performance and identify areas contributing most to revenue.</p><h4>2. Marketing Operations</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*c2VRpRUeD-a_qULEg13e3A@2x.jpeg" /></figure><p>The Marketing dashboard focuses on customer acquisition and conversion performance.</p><p>Key metrics include:</p><ul><li>Website Sessions</li><li>Product Views</li><li>Cart Additions</li><li>Purchases</li><li>Conversion Rate</li></ul><p>A funnel visualization illustrates customer progression from homepage visits to completed purchases, making it easy to identify where users drop off during the buying journey.</p><p>Additional charts compare traffic sources, browser performance, geographic distribution, and traffic trends over time to evaluate marketing effectiveness.</p><h4>4. Customer &amp; Growth</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*S0V7M9PhkG7oIF5bcSCQ8Q@2x.jpeg" /></figure><p>The Customer &amp; Growth dashboard focuses on customer acquisition, retention, and customer value.</p><p>It includes:</p><ul><li>Total Customers</li><li>Customer Lifetime Value (CLV)</li><li>Repeat Purchase Rate</li><li>Average Orders per Customer</li><li>Average Order Value</li></ul><p>Customers are grouped into High, Medium, and Low Value segments based on lifetime revenue.</p><p>The dashboard also includes customer acquisition trends, customer segmentation, revenue by acquisition source, geographic distribution, and cohort analysis to understand how customer value changes over time.</p><h4>Key Insights</h4><h4>1. Revenue Performance</h4><ul><li>The business generated <strong>$10.74M</strong> in total revenue and <strong>$5.57M</strong> in profit, resulting in a <strong>51.9% gross margin</strong> across <strong>124.6K orders</strong>.</li><li>The <strong>Average Order Value (AOV)</strong> is <strong>$86.20</strong>, indicating consistent customer spending per transaction.</li><li>Product sales are concentrated in a few categories, with <strong>Jeans</strong> emerging as the top revenue driver.</li><li>Brands such as <strong>7 For All Mankind</strong> and <strong>True Religion</strong> contribute significantly to overall sales.</li><li>From a marketing perspective, <strong>Search</strong> is the leading customer acquisition channel, generating the highest volume of traffic and purchases.</li></ul><h4>2. Marketing Funnel Performance</h4><ul><li>The website records a healthy <strong>26.6% conversion rate</strong>, but a <strong>58.1% cart abandonment rate</strong> which shows a major opportunity for improvement.</li><li>More than half of customers who add products to their cart leave before completing their purchase.</li><li>Performance remains almost identical across <strong>Chrome, Firefox, and Safari</strong>, with each browser showing similar conversion and abandonment rates. This suggests the issue is not browser-related but likely caused by friction within the checkout process, such as unexpected costs, lengthy checkout steps, or limited payment options.</li></ul><h4>3. Customer Growth and Value</h4><ul><li>The platform has <strong>79.9K paying customers</strong>, with <strong>62.6%</strong> making only one purchase and <strong>37.4%</strong> returning to buy again.</li><li>While the business continues to attract new customers, increasing repeat purchases shows a significant opportunity for long-term revenue growth.</li><li>Customer segmentation shows that <strong>54.4%</strong> of customers fall within the <strong>Medium Value</strong> segment (lifetime revenue between <strong>$100 and $499.99</strong>), while <strong>43.1%</strong> are <strong>Low Value</strong> customers. Although <strong>High Value</strong> customers account for only <strong>2.5%</strong> of the customer base, they contribute a disproportionately large share of total revenue, which shows the importance of retaining and expanding this segment.</li></ul><h4>Recommendations</h4><h4>1. Reduce Cart Abandonment</h4><ul><li>A <strong>58.1% cart abandonment rate</strong> shows the largest opportunity to improve conversions.</li><li>Since abandonment is consistent across all major browsers, efforts should focus on simplifying the checkout experience rather than resolving browser-specific issues.</li><li>Display shipping costs earlier, offer guest checkout, provide multiple payment options, and reduce the number of checkout steps to improve completion rates.</li></ul><h4>2. Strengthen Customer Retention</h4><ul><li>With nearly two-thirds of customers making only a single purchase, retention should become a key business priority.</li><li>Personalized post-purchase email campaigns, loyalty programs, product recommendations, and limited-time offers can encourage repeat purchases and increase customer lifetime value.</li></ul><h4>3. Grow the Medium Value Segment</h4><ul><li>More than half of the customer base belongs to the Medium Value segment, making it the most promising group for revenue growth.</li><li>Cross-selling complementary products, product bundles, and free shipping thresholds slightly above the current <strong>$86.20 AOV</strong> can encourage higher spending and move more customers into the High Value segment.</li></ul><h4>4. Optimize Marketing Spend</h4><ul><li>Search currently drives the highest customer acquisition.</li><li>Marketing investment should continue to prioritize high-performing channels while regularly evaluating conversion rates across all traffic sources to ensure spending is directed toward channels that generate both customers and revenue.</li></ul><h4>5. Focus on High-Performing Products</h4><ul><li>Top-performing categories and brands should receive continued marketing and inventory support to maintain sales momentum.</li><li>At the same time, lower-performing products should be reviewed for pricing, merchandising, or promotional improvements to improve their contribution to overall revenue.</li></ul><p>🔗 V<strong>iew the Live Interactive Dashboard </strong><a href="https://datastudio.google.com/reporting/2a047aca-4e52-49fa-85f2-1f62c50b0d67"><strong>Here</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6ff3f3e9b508" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Amazon Products Analysis: A Complete Data Analytics Case Study with POSTGRESQL and Power BI]]></title>
            <link>https://modupe-giwa.medium.com/amazon-products-analysis-a-complete-data-analytics-case-study-with-postgresql-and-power-bi-e76babd6d3b1?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/e76babd6d3b1</guid>
            <category><![CDATA[power-bi]]></category>
            <category><![CDATA[postgresql]]></category>
            <category><![CDATA[data-analytics]]></category>
            <category><![CDATA[sql]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Sat, 04 Jul 2026 17:32:12 GMT</pubDate>
            <atom:updated>2026-07-04T17:32:12.834Z</atom:updated>
            <content:encoded><![CDATA[<h3>Introduction</h3><p>In the world of e-commerce, data is king. Understanding customer sentiment, pricing strategies, and product popularity can make or break an online retail strategy. For this project, I analysed the <a href="https://www.kaggle.com/datasets/karkavelrajaj/amazon-sales-dataset/data"><strong>Amazon Sales Dataset</strong></a> from Kaggle, which contains data for over 1,400 Amazon products, including their ratings, pricing, discounts, and customer reviews.</p><h3>Objectives</h3><ol><li>To transform a messy CSV file into a highly structured database using <strong>PostgreSQL.</strong></li><li>To build an executive-ready, interactive dashboard in <strong>Power BI</strong> to uncover meaningful business insights.</li></ol><h3>The Stack &amp; Architecture</h3><p>To mirror a realistic enterprise data pipeline, I used a multi-tool approach:</p><ol><li><strong>Data Source:</strong> Raw CSV from Kaggle (1,465 rows and 16 columns).</li><li><strong>Storage &amp; ETL (Extract, Transform, Load):</strong> <strong>PostgreSQL (pgAdmin 4)</strong> for writing scripts to clean messy text string inputs, handle currency symbols, and perform schema modifications.</li><li><strong>Data Visualisation &amp; Analytics:</strong> <strong>Power BI Desktop</strong>, connecting directly to my clean database to build a 2-page dashboard.</li></ol><h3>Step 1: Data Cleaning in PostgreSQL</h3><p>The raw e-commerce dataset came with standard “dirty data” challenges: prices included commas, discount percentages had % signs, and massive review text blocks caused initial database imports to fail.</p><h4>The Fix:</h4><ol><li><strong>Flexible Staging Table:</strong> I created a database schema utilising TEXT data types for all columns to prevent value-too-long errors during the initial file import.</li><li><strong>Dropping Redundant Columns:</strong> I dropped columns like img_link and product_link, which were unnecessary for high-level aggregated data modelling.</li><li><strong>Regex Cleansing:</strong> Using PostgreSQL’s REGEXP_REPLACE, I stripped out all non-numeric symbols, converted percentages into decimals and securely transformed the column data types into NUMERIC and INTEGER.</li></ol><p>SQL</p><pre>-- Example of stripping currency and converting data types<br>UPDATE amazon_sales <br>SET <br>    discounted_price = REGEXP_REPLACE(discounted_price, &#39;[^0-9.]&#39;, &#39;&#39;, &#39;g&#39;),<br>    actual_price = REGEXP_REPLACE(actual_price, &#39;[^0-9.]&#39;, &#39;&#39;, &#39;g&#39;);</pre><pre>ALTER TABLE amazon_sales <br>    ALTER COLUMN discounted_price TYPE NUMERIC USING discounted_price::NUMERIC,<br>    ALTER COLUMN actual_price TYPE NUMERIC USING actual_price::NUMERIC;</pre><p>You can explore the complete PostgreSQL data cleaning pipeline and view generation scripts directly in the <a href="https://github.com/Modupegiwa/Amazon-Product-Data-Analysis/tree/main">GitHub Repository</a>.</p><h3>Step 2: Dashboard Design &amp; Power BI Implementation</h3><p>Once the database was clean, I established an Import connection from Power BI to PostgreSQL. To make the dashboard highly user-friendly and structured, I designed 2 pages linked together with a unified top-navigation bar utilising interactive buttons with dynamic <strong>on-hover</strong> visual effects.</p><h4>Key DAX Metrics Formulated:</h4><ol><li><strong>Total Products Count:</strong> Total Products = DISTINCTCOUNT(amazon_sales[product_id])</li><li><strong>Product Tiering:</strong> Categorised items automatically into market segments using a SWITCH conditional column based on price thresholds.</li></ol><figure><img alt="Page 1: Executive Summary" src="https://cdn-images-1.medium.com/max/585/1*AVSSb3UY9W4D48th8p5oPA@2x.jpeg" /></figure><figure><img alt="Page 2: Product Performance and Review Analysis" src="https://cdn-images-1.medium.com/max/585/1*acmj4lDU9veK1ksQsQQelw@2x.jpeg" /></figure><h3>Step 3: Data Insights &amp; Dashboard Breakdown</h3><p>Based on the completed dashboard views, several prominent trends stand out regarding pricing models and category dominance:</p><h4>1. High-Level Performance Overview</h4><ul><li><strong>The Scale:</strong> The final database holds a clean footprint of <strong>1.4K Total Products</strong> spanning across multiple consumer segments.</li><li><strong>Healthy Customer Sentiment:</strong> The ecosystem maintains a remarkably high <strong>Average Rating of 4.1 out of 5.0</strong>, representing broad consumer satisfaction across the catalogue.</li><li><strong>Aggressive Discount Strategy:</strong> The average percentage discount across the entire site sits at a staggering <strong>47.7%</strong>. This reveals that Amazon listings rely heavily on psychological “anchor pricing” (showing a high actual price but selling at a near-permanent markdown) to incentivise checkouts.</li><li><strong>Volume Dominance:</strong> Total customer engagement is massive, capturing <strong>26.8 Million Total Customer Reviews</strong> across the dataset.</li></ul><h4>2. Category and Inventory Distribution</h4><ul><li><strong>The Catalogue Heavyweights:</strong> Inventory concentration is heavily skewed toward <strong>Electronics</strong> and <strong>Computers &amp; Accessories</strong>.</li><li>Looking at the product counts by category, <strong>Electronics</strong> leads the volume with over 500 items, closely followed by <strong>Computers &amp; Accessories</strong> (approaching 400 items). Meanwhile, niches like <em>Office Products</em>, <em>Home Improvement</em>, and <em>Musical Instruments</em> make up tiny fractional pieces of the total inventory.</li><li><strong>Electronics</strong> captured an impressive <strong>57.80K review engagements</strong> compared to <strong>20.39K</strong> for Computers &amp; Accessories.</li></ul><h4>3. Top-Tier Highlights &amp; Sentiment Distribution</h4><ul><li>When focusing on the highest-rated inventory tier, products like the <em>Amazon Basics Wireless Mouse</em> and specialised <em>MFi Certified USB-C to Lightning Cables</em> consistently top the charts with <strong>5.0 ratings</strong>.</li><li>Interestingly, while the peak of the rating histogram sits comfortably in the 4.0–4.5 star range, products with lower ratings still see substantial discount rates, indicating that sellers try to move less-popular stock by intensifying discount rates (climbing up to <strong>94.0%</strong> in specific clearance sub-categories).</li></ul><h3>Key Takeaways &amp; Conclusion</h3><p><strong>E-commerce success on this marketplace relies heavily on aggressive discounting.</strong> With a baseline average markdown of nearly 48%, competitive pricing isn’t just an advantage — it’s an entry requirement, particularly in high-volume categories like Electronics</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e76babd6d3b1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2023: Year In Review]]></title>
            <link>https://modupe-giwa.medium.com/2023-year-in-review-b49a263955d4?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/b49a263955d4</guid>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Sat, 30 Dec 2023 08:35:53 GMT</pubDate>
            <atom:updated>2023-12-30T08:35:53.596Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GhfQb1ZBFhjuPF7XhD2PpA.png" /></figure><p>As 2023 draws to a close, I take a moment to reflect on the highlights of the year which include struggle, success, and self-discovery.</p><p>My journey was shaped by volunteering, interning, academic achievements, and self-care practices.</p><p>2023 wasn’t just a year; it was my journey through a series of relationships. It was a celebration of progress, connections, and the beauty found in relationships.</p><h3>Volunteering</h3><p>In the whirlwind of activities that defined my year, volunteering played a crucial role in shaping my sense of community and social responsibility.</p><h4>TESA Inaugural Conference</h4><p>In June, I Served as a member of the welfare committee and the team lead for the Registration and Correspondence Team for the TESA Inaugural Conference with Kehinde Yusuf, the Wood Products Engineering Department President. It was a challenging yet fulfilling experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6vh2iVn776Yo2-bq6Uhm_w@2x.jpeg" /></figure><p>Special thanks to the Conference Commitee Lead, Oluwatosin Ogundinmu for allowing me to lead and volunteer.</p><h4><strong>University of Ibadan Research and Development Hackathon</strong></h4><p>In August, I got selected as one of the organizers for the University of Ibadan Research and Development Hackathon which is organized by the University of Ibadan Research and Development Fair to provide a platform for students with diverse skills and backgrounds to come together to collaboratively create innovative solutions to address pressing challenges on campus and beyond.</p><p>The event held on October 31 at Trenchard Hall, University of Ibadan, provided a platform for aspiring researchers and developers and showcased the potential of harnessing collective intelligence for impactful solutions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V9HkuiL0RBTWN9TQszXI7Q@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6x6eZ32A2pLaN0sPu3v9pQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kl1LRraotgE-SQUmi3yXOw@2x.jpeg" /></figure><p>To ensure the event was a success, I worked with individuals such as Habeeb Lanlehin, Victoria Ayodeji, Emmanuel Aiyede, Obaloluwa Animashaun, <a href="https://medium.com/u/9731fd700c7a">Oluwatosin Ogundimu</a>, <a href="https://medium.com/u/2de0e9bd839b">Badejo Prosper</a>, Tiolu, Moyo, and others.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JOw2POvrYI3emfyUhx_A0w@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5zOn80jo5D2eu0HLTtf9YQ@2x.jpeg" /></figure><h4><strong>WAAW Foundation</strong></h4><p>My experience in volunteering took on a new level when I became a fellow and volunteer for the WAAW Foundation, University of Ibadan, in August. The WAAW UI team organized multiple outreaches for secondary school students in Ibadan and I had an amazing experience volunteering at Methodist Grammar School.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SzGcTDkkygG8c2PX6p5O6g@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KoaGFSGnjxjGvFT93JBJDQ@2x.jpeg" /><figcaption>WAAW UI Fellows</figcaption></figure><p>It is always nice working with the WAAW UI lead, Chidinma, Co-lead, Kelechi, the fellows’ manager, Young, and other fellows.</p><h4><strong>IEEE Women in Engineering</strong></h4><p>During the 2021–2022 session at the University of Ibadan, I was elected as an executive of the IEEE Women in Engineering UI student branch, which enabled me to participate in the organization’s goal of empowering women in technology and inspiring and encouraging them on their tech journeys.</p><p>The Technical Skill Initiative with the theme “Navigating Through Tech: Career Development in Tech &amp; Interview Preparation for Your Dream Job” was one of the events organized throughout the academic session. The LPI Innovation Hub hosted the event in April 2023, and it included an incredible speaker lineup that included Paloma Peter-Dhozo, Favour Olusoji, Adedigba Oluwasegun, and others.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0LAOi9uH5BLeiBQ7cXU6Vw@2x.jpeg" /><figcaption>WIE UI Executives</figcaption></figure><p>You can learn more about the event here: <br><a href="https://medium.com/wie-uisb/wie-uisb-tech-skills-initiative-90a5b115c3ef">https://medium.com/wie-uisb/wie-uisb-tech-skills-initiative-90a5b115c3ef</a></p><p>I am grateful to Temitope Ajayi, the WIE Chair, for giving me the chance to volunteer, and I had fun working with the other executives from <a href="https://medium.com/u/5a649d153deb">Agangan Comfort</a>, Adebanke, Damilola, and others.</p><h4><strong>GDSC TechQuest Event</strong></h4><p>I volunteered for the Google Developers Students Club (GDSC) UI TechQuest Event in November. The event was an excellent platform for sharing knowledge, networking, and inspiring tech enthusiasts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ko4T1kuet9sgIJSlL-8S-w@2x.jpeg" /></figure><p>As a volunteer, I assisted with registration, Signing in attendees, and distribution of event materials.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_2Yi2HXI_YPCEu9uJ9YLXA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eljxBRGXTY-HlE8DWK7hNA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-piQZj8kjUwLqjQEzTr4Sg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ny6m594HFhnyWAdGwTEDgw@2x.jpeg" /></figure><p>I am grateful to the GDSC team lead, Bukola Olopade, and the Co-leader, Eniolaoluwa Bakare for the opportunity. It was an amazing experience working with the GDSC team!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*L4krQauq_hXbzzv_txEU8A@2x.jpeg" /></figure><h3>Internship</h3><p>I completed my three-month internship at the Innovation Hub and Design Studio at the University of Ibadan. The internship was a transformative journey into the world of practical application of engineering skills.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*X52_BsvRhZC2aJYf0Zy1tg@2x.jpeg" /></figure><p>I engaged in Arduino-based projects, learned how to design, and model 3D objects using Autodesk Fusion 360, used slicing programs like Cura and PrusaSlicer to maximize print quality, and operated 3D printers to print 3D designs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XauE9h4My8WXRYxCr2wVKQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*GTjNLJzWsdXlh00Vd-Lc4w@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_ReJUIgBf12enLIKAOOO6A@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*RRJY9D8iK9AUbQgBFCgANA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6ZxjSZ6DDqXzolho44cV6w@2x.jpeg" /></figure><p>During my internship, I also learned how to operate a CNC milling machine and laser cutter. Additionally, I learned how to use the Python programming language for data visualization and analysis.</p><p>My favorite part of the internship was working in a group to cut out a desk organizer design using a laser cutter. To get the best results, we used the Laser Works software to fine-tune the cutting settings, engrave a name, and adjust parameters like speed and power. We used the laser cutter to accurately cut the components after adjusting the settings. The components were skillfully assembled to create the completed desk organizer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PWxMtd-xUKwQY5f442zOGg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uDsWc8HRak0hjY161nbghw@2x.jpeg" /></figure><p>I enjoyed working on this project and was impressed with the level of precision achieved.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Rbo7f65zWjnN_EqfDP-fEw@2x.jpeg" /></figure><p>I am grateful to Engr. Olusola, Engr. Mayowa, UI Design Studio staff, and my fellow interns for making my time there great.</p><h3>School: Academic Excellence and Leadership</h3><p>Academically, the year was marked by achievements and leadership roles.</p><p>I ended the last academic session with a 3.80 GPA and significantly increased my CGPA. The academic challenges were met with determination, hard work, and plenty of TDBs.</p><p>I ran for the post of General Secretary for the Industrial Engineering Students Association (IESA). I had to go through the election process during the academic session which comprised the Electoral screening, Press screening, and Manifesto and I got elected. I am grateful to my classmates, friends, and IEASites for supporting me and voting for me.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BS_h2qYg6MB9fqi_VnDAag@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7WErR-flAlDKXFIYcvVkkw@2x.jpeg" /></figure><p>In addition, I was elected Vice President of the National Society of Black Engineers, University of Ibadan Branch. Our first webinar on “Securing Global Internships for Undergraduates” was arranged by the NSBE UI team, and it was a rewarding experience. We invited James Olayinka, Prince Ikpabi, and Pelumi Egunjobi to speak about their internship experiences and give other students insightful information and chances. I am grateful to the NSBE President, Tumininu, Program Chair, Ejiro, the Financial Secretary and the event host, Emmanuella, the speakers, and other executives for making the event a success.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xRu3jWIo59KhAbR1LkYCfw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/540/1*rpgxi-er5k6ZLs6BzeiDgg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*R4dGOK3zB1S13HOaLrLfYg@2x.jpeg" /></figure><h3>Self Care</h3><p>Beyond the academic and professional areas, self-care became a priority.</p><p>I experimented with a variety of hairstyles this year, including Fulani styles — what my friend Stephen refers to as “half-done hair” — and braids.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*qzM66W6QvEUt04RFOFYFCg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WRpA6k25E6vODAlJE7JlzQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*educOKZDu4qeH3cGKgK7rA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/1*zUJ5pnkoiKrEXznlmkpkvw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*_w8gaRTKvzT7qGUe402EXQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*LOLBCZm69pjgmrJ6d3hJFA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-k5JaAZ8gvIKFu58lwDcaw@2x.jpeg" /></figure><p>I also committed to getting at least six hours of sleep every day. Although maintaining the consistency wasn’t easy, I’m glad I managed to do it.</p><p>And as November unfolded, so did a newfound tradition — on November 21st, I had a great dinner with friends to celebrate my birthday.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/354/1*qngAohOvAYwTRqRnwmWoUw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*m6GsVOYxs82LAJiiY6xZoA@2x.jpeg" /></figure><p>It was a moment of joy, reflection, and connection — a fitting ending to a year of growth, self-love, and meaningful experiences.</p><h3>Relationships: Building Connections and Making Memories</h3><p>The year was not only about professional and personal growth but also about building meaningful connections.</p><p>I participated in intellectual discussions, went to several conferences and events, and made new friends who widened my social circle. I attended several events, including the popular Ojude Oba festival, the Women TechMakers Ibadan IWD event, the Google I/O Extended Ibadan event, and the Ibadan Tech Expo event.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/414/1*-yQ_0k4oMVNBipX7So7xwg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8-u2rs6MYkSjooW33ehGNw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GCBsQIEUTRQgcunYPJZSUg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*WtSsJImqMcPe49CkLTHSCA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EzmRrKadPaV1DJAx1ckHaQ@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6morZGi_DK90lQ9TuJsWwg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*g-1y0ZRIbKdtMRWZ-4JJcg@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FdJmQCFPr4_iAIjGeJRUUw@2x.jpeg" /></figure><p>I also went to events like my department’s dinner party, picnic, and my cousin’s wedding in Lagos.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/504/1*-z_bLYxJvE4p4uVpwwv8Nw@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pEOuN8-zqIYVR2Hf5jG9PA@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*WQ7VPb0uKAhLuvP2azUz8Q@2x.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KdpaOBVCO_U1ospdfUMP8w@2x.jpeg" /></figure><p>Navigating through conflicts with resilience and maturity added emotional intelligence to my growth journey.</p><h3>Some Things Didnt Go As Planned In 2023</h3><p>In the face of achievements, 2023 also presented its share of challenges:</p><ul><li>I Received multiple rejection emails.</li><li>I couldn’t complete the Udacity Nanodegree Program.</li><li>I failed to get into Junior Chambers International (JCI) University of Ibadan Chapter.</li><li>I couldn’t reach my savings goal.</li></ul><h3>Summary</h3><p>2023 was a year of growth, resilience, and profound self-discovery. The combination of professional achievements, self-care, and meaningful relationships created experiences that will undoubtedly shape my trajectory in the years to come.</p><p>I want to start by giving thanks to God for the year; for keeping me safe, guiding my decisions, and providing me with advisors to help me through the tough decisions that come with life. I am grateful to God for keeping everyone close to me safe and loved.</p><p>This year, having faith was the most important lesson I learned. I still recall giving my mother a call and telling her about my worries. When I tell her I’ve given up, she won’t stop encouraging me to have faith and not give up. I sincerely appreciate my family’s encouragement, prayers, and support.</p><p>I am grateful to my friends. The love was simply too much. Because of you guys, I realized the value of having good friends.</p><p>I also want to thank everyone who helped make 2023 an incredible year for me. Thank you for believing in me.</p><p>As I bid farewell to 2023, I carry forward the lessons learned, the connections made, and the resilience developed, prepared to welcome the unknown adventures that lie ahead in the future chapters.</p><h3>Thank you for reading!</h3><blockquote>I wish everyone a Happy New Year. May this new year bring new happiness, new goals, new achievements, and a lot of new inspirations in your life. Wishing you a year fully loaded with happiness!</blockquote><h3><strong>Connect with me:</strong></h3><p>https://linktr.ee/modupe_giwa</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b49a263955d4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploratory Data Analysis On Hotel Booking]]></title>
            <link>https://modupe-giwa.medium.com/exploratory-data-analysis-on-hotel-booking-77be7adf48cd?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/77be7adf48cd</guid>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-visualization]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Fri, 23 Sep 2022 21:32:29 GMT</pubDate>
            <atom:updated>2022-09-23T21:32:29.443Z</atom:updated>
            <content:encoded><![CDATA[<h3>Dataset</h3><p>The <a href="https://www.kaggle.com/datasets/jessemostipak/hotel-booking-demand">Hotel booking demand</a> data set contains booking information for a city hotel and a resort hotel, and includes information such as when the booking was made, length of stay, the number of weekends or weeknights the guests stayed and the number of available car parking spaces, among other things.</p><p>All personally identifying information has been removed from the data.<br>The data set contains 119390 guests with 32 feature sets.</p><h3>Outline:</h3><p>- Import Packages</p><p>- Extract And Read Data</p><p>- Get Information About The Data</p><p>- Data Preparation: <br> 1. Find Percentage Of Missing Values <br> 2. Converting Column Values <br> 3. Changing Datatypes <br> 4. Checking Duplicate <br> 5. Find The Total Nights Stayed By Guests</p><p>- Descriptive Statistics</p><p>- EDA</p><p>- Relationships</p><p>- Drop Some Features</p><p>- Correlation: <br> 1. Correlation Matrix <br> 2. Heat Map</p><h4>Import Packages</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1006/1*dWIM6dkrDBzdVDsiDpaWhw.png" /></figure><h4>Extract And Read Data</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7OOytqJerlRnTMeQWudj7w.png" /></figure><h4>Get Information About The Data</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/462/1*1KKrMtjvUCiDD7AXCuUYKg.png" /></figure><h4>Data Preparation</h4><ol><li>Find the Percentage Of Missing values</li></ol><ul><li>There are missing values in the “agent”, “company” and “children” columns.</li><li>Fill in the missing values in the “agent” and “company” columns.</li><li>Remove rows with missing values in the “children” column.</li><li>Fill the “NaN” value in the “children” column using the mode.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/1*S6gP2KfFpxKNN9WXuKUrTw.png" /></figure><p>2. Converting Column Values</p><ul><li>Convert “Undefined” to “TA/TO” in the “distribution_channel” column</li><li>Convert Values Of The “meal” column</li><li>Convert Values Of the “is_canceled” Column to “Cancelled” and “Not Cancelled” from 0 and 1</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uHvIg5LGOr9mffAHBKDi7g.png" /></figure><p>3. Changing Datatypes<br>- Convert datatypes of “children”, “agent”, and “company” columns to integer<br>- Convert the datatype of the “reservation_status_date” column from object to DateTime</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2N7vqCEFGOnS5uR5N-Gt3A.png" /></figure><p>4. Checking Duplicates</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/758/1*hG-90Shkep1dWztrkucwVQ.png" /></figure><p>5. Find the Total Nights Stayed By Guests</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/840/1*HoWHqSXyvpFN3T8S6ALmbw.png" /></figure><h4>Descriptive Statistics</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/566/1*xlPGgXWsbwEpjgrVYgcOKg.png" /></figure><h4>EDA</h4><p><strong>Hotels</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/656/1*1sj-qGb5eY69bS9KplaRiw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*JjDX0D8pSkVD17a2pAernw.png" /></figure><p>More than 60% of the hotels booked are City Hotels</p><p><strong>Market Segment</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*clM7hpmddatrNjSw-3ExRA.png" /></figure><p>More than 50% of the bookings were done through Online travel agents.</p><p><strong>Distribution Channels</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7f57DfjFrZXp3lg2tdhuvQ.png" /></figure><p>More than 80% of the bookings were done through travel agents/operators.</p><p><strong>Meals</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pDl6xlEDa6_63WuzGUwtww.png" /></figure><p>More than 70% of the guests book breakfast and almost 90% of the total guests book meals.</p><p><strong>Customer Type</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SZ-FBWhXhsmHFjA4DCm4aw.png" /></figure><p>The majority of the bookings are not part of a group or contract.</p><p><strong>Deposit Type</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*De4ES-N0OfSh3MzIIn33gA.png" /></figure><p>The majority of the guests made no deposit</p><p><strong>Type Of Guest</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jGVN216AanStJAXwYlG9xQ.png" /></figure><p>Almost all of the guests are new guests.</p><p><strong>Reserved Room Types</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uzSCNTPw5AJqOL9d0RN3gA.png" /></figure><p>Most of the guests reserved room A, a few reserved rooms D &amp; E and the demand for the rest are minimal.</p><p><strong>Assigned Room Types</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lUP8UFFZ2D_aHacipZ8P3g.png" /></figure><p>Most of the guests are assigned room A, a few were assigned rooms D &amp; E and the rests are minimal.</p><p><strong>Reservation Status</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ErIHXdrqLL9HQgXcnjZQMQ.png" /></figure><p>Most of the guests checked in and already departed.</p><p><strong>Total Night Stayed</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*F19ZfXr7_qANodS00hmEiw.png" /></figure><p>The most popular stay duration for the hotels is three nights.</p><p><strong>Total Night Stayed In Each Hotel</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kP1JlV5bb9eYJ5Oluk59JQ.png" /></figure><p>The most popular stay duration in Resort Hotel is one, seven, two, three and four nights. The most popular stay duration in City Hotel is three, two, one and four nights.</p><p><strong>Top 20 Countries</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IF0wS2LWB5UlppOeZcbFFA.png" /></figure><p>Portugal has a higher number of bookings than any other country.</p><p><strong>Lead Time</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZDspjg2bwKWRGhsuhzOvIw.png" /></figure><p>Most bookings are done within 100 days before checking in the hotel.</p><h4>Relationships</h4><p><strong>Number Of Bookings Cancelled By Guests</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vt1ZqbCkDXDBCBopAdP3yw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8gKXjFysIU7FHPIejGvEnw.png" /></figure><p>Guests in Resort hotels cancel their bookings less often than guests in City hotels.</p><p><strong>Number Of Bookings Cancelled By Type Of Guests(Repeated/New Guests)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MVuuVTkKBPlLAgEwxt0g1g.png" /></figure><p>Repeated guests cancel their bookings less than new guests.</p><p><strong>Number Of Bookings Cancelled By Market Segments</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lg8cf06JkIv1Hh_Qw5st3g.png" /></figure><p>Online travel agents, offline travel agents/operators and the direct segment have high cancellation rates compared to the rest.</p><p><strong>Number Of Bookings Per Year</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Cu8VKNet6CWvYCHlIksskg.png" /></figure><p>Both resort and city hotels had their highest bookings in 2016. City hotels have higher bookings in 2017 compared to resort hotels. Both had almost the same number of bookings in 2015.</p><p><strong>Number Of Bookings Per Month</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vSf_Jw7auzJI8KOrHdrwkg.png" /></figure><p>November, December, January and February are the months having the least books. July and August are the months having the highest bookings.</p><p><strong>Number Of Bookings Per Customer Type</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_SR758RkHp7HrWbH-3FYIA.png" /></figure><p>Transient and Transient-party guests mostly book City hotels while contract and group guests almost have the same number of bookings in resort and city hotels.</p><p><strong>Bookings Per Required Car Space</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8WLgj0SKk-D5ZeO9Iqodkw.png" /></figure><p>The majority of the guests don’t need car parking spaces while few guests need parking spaces.</p><p><strong>Average Number Of Special Requests Made By Top 20 Countries</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u8elV60uGJaqRu_IgZQ_tg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DHsjti_A6VW92flZnm2mWg.png" /></figure><p>Botswana has the highest number of special requests among the countries.</p><p><strong>Average Number Of Special Requests Made By Customer Type</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MHtzfxf9sXGlABXI-KLvQg.png" /></figure><p>Contract guests have the highest number of special requests while transient-party guests have the lowest number of special requests.</p><p><strong>Average Number Of Special Requests Made Per Guests Across Months</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4sSKbjmXHSE26GoYPh3z6A.png" /></figure><p>An almost similar number of special requests were made by guests across months but slightly more were made in August, July and December.</p><p><strong>Average Daily Rate Per Hotel</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Dk15hxEe90mQdItK9tKVog.png" /></figure><p>The average daily rate of Resort hotels is more spread than City hotels.</p><p><strong>Average Daily Rate Per Month</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DqrBl1gv6zZofakphHjGrg.png" /></figure><p>The average daily rate for both hotels is higher mid-year. City hotels have higher daily rate beginning and the start of the year compared to Resort hotels.</p><h4>Drop Some Features</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1006/1*dfk6SgQ1wve_nCTimF8AGg.png" /></figure><h4>Correlation</h4><p><strong>Correlation Matrix</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FH9J3MBpwN5Tfe7iIUeTcQ.png" /></figure><p><strong>Heat Map</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/907/1*vjFQbjykhJ54snJhuka9xw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aJToy-Ar1ERdLbuT3tXz1w.png" /></figure><h4>Check Out This Project On GitHub:</h4><p><a href="https://github.com/Modupegiwa/Hotel_Booking.git">https://github.com/Modupegiwa/Hotel_Booking.git</a></p><h4>Contact:</h4><ul><li><a href="https://twitter.com/modupe_giwa">Twitter</a></li><li><a href="https://www.linkedin.com/in/modupe-giwa-7b9a7417b/">LinkedIn</a></li><li><a href="https://github.com/Modupegiwa">GitHub</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=77be7adf48cd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Exploratory Data Analysis On Heart Disease Data Set]]></title>
            <link>https://modupe-giwa.medium.com/exploratory-data-analysis-on-heart-disease-data-set-69236bdf2cce?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/69236bdf2cce</guid>
            <category><![CDATA[health]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[heart-disease]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Sat, 17 Sep 2022 01:11:50 GMT</pubDate>
            <atom:updated>2022-09-17T01:11:50.313Z</atom:updated>
            <content:encoded><![CDATA[<h3><strong>Introduction</strong></h3><p>Cardiovascular Diseases(CVDs) are the number one cause of death globally, taking an estimated 17.9 million lives each year, which accounts for 31% of all deaths worldwide.<br>Heart failure is a common event caused by CVDs and this dataset contains 14 features that can be used to predict mortality by heart failure.<br>Most cardiovascular diseases can be prevented by addressing behavioral risk factors such as tobacco use, unhealthy diet, obesity, physical inactivity, and harmful use of alcohol using population-wide strategies.<br>Early prediction of cardiac diseases can decrease the mortality rate and overall complications.</p><h3>Dataset</h3><p>The dataset used in this project is <a href="https://www.kaggle.com/datasets/johnsmith88/heart-disease-dataset">Public Heart Dataset</a>. The dataset contains features that can be used for predicting heart disease.<br>It consists of 1025 patients with 14 features set:<br>1. age (Age in years)<br>2. sex: (1= male, 0= female)<br>3. cp (Chest Pain Type): [0= Asymptomatic, 1= Atypical angina, 2= Non-anginal pain, 3= Typical angina]<br>4. trestbps (Resting Blood Pressure in mm/hg)<br>5. chol (Serum Cholesterol in mg/dl)<br>6. fbs (Fasting Blood Sugar &gt; 120mg/dl): [0= False, 1= True]<br>7. restecg (Resting Electrocardiographic Results): [0: showing probable or definite left ventricular hypertrophy by <br>Estes’criteria, 1: normal, 2: having ST-T wave abnormality]<br>8. thalach (Maximum Heart Rate)<br>9. exang (Exercise Induced Angina): [1= True, 0= False]<br>10. oldpeak (ST depression induced by exercise relative to rest)<br>11. slope (Slope of the peak exercise ST segment): [0: downsloping, 1: flat, 2: upsloping]<br>12. ca (Number of major vessels): [0–3]<br>13. thal (Number of thal): [1= normal, 2= fixed, 3= reversible defect]<br>14. target (Heart disease status): [0= disease, 1= no disease]</p><h3>Outline For The EDA</h3><p>1. Import and get information about the data set<br>2. Cleaning the data<br>3. Data exploration for binary and categorial variable distribution<br>4. Data exploration for continuous variable distribution<br>5. Relationship between heart disease and sex(Gender), cp(Chest pain type), fbs(Fasting blood sugar), and ca(Number of major vessels)<br>6. Correlation and Heat map</p><h4>1. Import and get information about the data set</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/640/1*lFRNOREK8yGikafyujv6yQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/931/1*mnQnuc187Bn0h-OGkMP6TQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/757/1*i7vDW7xLvbzX-o23KS-YtQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/952/1*pwv5OneztvZ3td20PsmYAg.png" /></figure><p>There are 1025 rows with 14 variables</p><h4>2. Cleaning the Data</h4><p><strong>a. Check for null values</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*o5UAq21bj89s6xvLl1oe4g.png" /></figure><p>There are no null values</p><p><strong>b. Look for mistakes in the values of the variables</strong><br>1. The values of the column “ca” range from “0- 4” in the data instead of “0–3”. Find the value “4” and change it to NaN</p><p>2. The values of column “thal” ranges from “0–3” in the data instead of “1–3”. Find the value “0” and change it to NaN</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*5ijUiHH6AgXJao-FtsDODw.png" /></figure><h4>Data Exploration</h4><p><strong>1. Gender Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/661/1*_lFa2SxD6Gat9PiuIg_FKA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R0zwItxvwYZUvczMfrE_kw.jpeg" /></figure><p>There are more male patients than female patients</p><p><strong>2. Chest Pain Type(cp) Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cCCH2rvaFoSkc8zclKccXA.jpeg" /></figure><p>There are more patients that have Asymptomatic chest pain.</p><h4>3. Fasting Blood Sugar(&gt;120mg/dl)</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hm9Canc7EfMl8yLqUzfVKQ.jpeg" /></figure><p>Fasting blood sugar is an indicator for diabetes(if fbs &gt; 120mg/dl). There are more patients who aren’t diabetic.</p><p><strong>4. Resting Electrocardiographic Distribution(restecg)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MREEk-2vkEg_S5RruVYcRg.jpeg" /></figure><p><strong>5. Exercise-Induced Angina Distribution(exang)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yc_LwnZdl8XHYQZ4TEBTOA.jpeg" /></figure><p><strong>6. Slope Of The Peak Distribution(slope)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZI1nJUeJ9rDAJtKBKpdn4Q.jpeg" /></figure><p><strong>7. Number Of Major Vessels Distribution(ca)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2N8AhM3KzCdUok3SHQxB8g.jpeg" /></figure><p>There are more patients with 0 major vessels.</p><p><strong>8. Thal Distribution(thal)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MTJhAb2xAVC1FTQPjLjkcQ.jpeg" /></figure><p>There are more patients with fixed defects.</p><p><strong>9. Heart Disease Status Distribution(target)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1R5B2exbtF96gTq9VVswmA.jpeg" /></figure><p>There are more diseased patients compared to non-disease patients.</p><h3>Descriptive Statistics</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/684/1*utbmoW-s3B6uVfRpJ88Sow.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/936/1*tRa5nuEgHk193c-JaiVNEg.png" /></figure><h3>Continuous Distribution</h3><p><strong>10. Age Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/646/1*MFmP6PMMLn3GAuJJTDGlcw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cq0qinUWOfZ5k_LrPtT2UA.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/612/1*CldaxY_KWZ-_AoqdASOKTQ.png" /></figure><p>From the chart:<br>a. The value of the skewness is negative.<br>b. The mean &lt; median.<br>It shows that the distribution is skewed left.</p><p><strong>11. Resting Blood Pressure(trestbps) Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*e8PfVxaP78j4KM_At7iz1w.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/697/1*fmjsFLXQw7XA44BgkuYmgA.png" /></figure><p>From the chart:<br>a. The value of the skewness is positive.<br>b. The mean &gt; median.<br>It shows that the distribution is skewed right.</p><p><strong>12. Serum Cholesterol(chol) Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4VKGbInpQXv5yod1ke6uCw.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/1*jhsWCBNqHyz0nTL75ClV8Q.png" /></figure><p>From the chart:<br>a. The value of the skewness is positive.<br>b. The mean &gt; median.<br>It shows that the distribution is skewed right.</p><p><strong>13. Maximum Heart Rate(thalach) Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d9ja51lnS68-JIIoi3gk9Q.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/624/1*Z0N8UlcZ1cn8h5xlEkHj3g.png" /></figure><p>From the chart:<br>a. The value of the skewness is negative.<br>b. The mean &lt; median.<br>It shows that the disribution is skewed left.</p><p><strong>14. Old Peak Distribution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1Ukv2zx5BsvVH5OS4GxK8g.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/1*8SYF998mF650QBhfYJGqsg.png" /></figure><p>From the chart:<br>a. The value of the skewness is positive.<br>b. The mean &gt; median.<br>It shows that the distribution is skewed right.</p><h3>Distribution And Relationship</h3><p><strong>1. Heart Disease(target) Distribution Based On Gender(Sex)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*3F-zVVxns9OVd6qq6GJqoA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*canYp5822RjKnA5_5ynonw.jpeg" /></figure><p>From the bar graph, It is shown that males tend to have heart disease compared to females.</p><p><strong>2. Heart Disease(target) Distribution Based On Number Of Major Vessels(ca)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zWw9wCLrkzDMQjSy9USnvQ.jpeg" /></figure><p>From the bar graph, Patients with 0 major vessels tend to have heart disease while Patients who have a number of vessels between 1 to 3 tend not to have heart disease.</p><p><strong>3. Heart Disease(target) Distribution Based On Chest Pain Type(cp)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rSFvxBmwKGfsE3EkQ90O1A.jpeg" /></figure><p>From the bar graph, Most patients having typical angina, atypical angina and non-angina chest pain tend to have heart disease while patients with asymptomatic chest pain tend not to have heart disease.</p><p><strong>4. Heart Disease(target) Distribution Based On Fasting Blood Sugar(fbs)</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z9Tbj4kqyIH8TVCSNV0RkQ.jpeg" /></figure><p>Fasting blood sugar is an indicator for diabetes(if fbs &gt; 120mg/dl). From the bar graph, There are a higher number of heart disease patients who isn’t diabetic. This shows that fbs might not be a strong feature in showing the relationship between heart disease patients and non-disease patients.</p><h3>Correlation</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/685/1*Ag5VRHx6yubuT0TYXG7Zxg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*HK-eiGNqZsIwcUIkOnUVYg.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/927/1*s5ljUBZoipdhm2ITmIrMsA.png" /></figure><p>a. Maximum heart rate (“thalach”) and slope of the peak exercise(“slope”) shows a good positive correlation with Heart disease status(“target”)<br>b. “sex” and “oldpeak” shows a good negative correlation with heart disease status(“target”)<br>c. Serum cholesterol(“chol”) and resting blood pressure (“trestbps”) show a low correlation with heart disease status (“target”).</p><h4>Check out this project on GitHub:</h4><p><a href="https://github.com/Modupegiwa/Heart-Faliure-Prediction">https://github.com/Modupegiwa/Heart-Faliure-Prediction</a></p><h3>Connect With Me On:</h3><ul><li><a href="https://twitter.com/modupe_giwa">Twitter</a></li><li><a href="https://www.linkedin.com/in/modupe-giwa-7b9a7417b/">LinkedIn</a></li><li><a href="https://github.com/Modupegiwa">GitHub</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=69236bdf2cce" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fun Facts About Me]]></title>
            <link>https://modupe-giwa.medium.com/fun-facts-about-me-71c4d721181a?source=rss-189c00762943------2</link>
            <guid isPermaLink="false">https://medium.com/p/71c4d721181a</guid>
            <dc:creator><![CDATA[Modupe Giwa]]></dc:creator>
            <pubDate>Fri, 08 Apr 2022 06:03:19 GMT</pubDate>
            <atom:updated>2022-04-08T06:03:19.496Z</atom:updated>
            <content:encoded><![CDATA[<p>I have been thinking about writing for a while now and I couldn’t think of anything. I realized that this is an opportunity for me to write about myself.</p><p>I spent an hour thinking about fun facts about me and I was able to gather some things.</p><p>So here we go:</p><ul><li>I write letters to my future self through a website called &quot;future me&quot;</li><li>I can’t ride a bicycle.</li><li>I am scared of many things. Like water, crossing road, insects, snakes, etc.</li><li>I am very introverted, I don’t like going out or meeting people. I think I enjoy my company too much.</li><li>I watch movies and series a lot. I can spend a whole day on my bed watching movies.</li><li>When I was 8, I slipped in front of a bucket of hot water and I burnt my right hand.</li><li>I love sleeping, to me that’s the only thing that helps me escape from reality.</li><li>I mostly sleep during the day. I am very active in the night.</li><li>I eat onions a lot because I believe it will help my bad eyesight (It is still bad tho).</li><li>I love reading books and listening to music.</li><li>I mostly listen to pop, hip-hop, R&amp;B and rap songs.</li><li>I am a big fan of Davido and BTS.</li><li>I have a journal. It helps me think better.</li><li>I was very good at mathematics till I entered University.</li><li>I just have this dislike for biology maybe because I find it boring.</li><li>I don’t know how to receive compliments. I believe I don’t deserve it or they are just too much.</li><li>Everytime I watch a sad movie, I cry.</li><li>I feel really bad when someone is not comfortable around me.</li><li>I have this bad habit of giving up or running away when things become too hard.</li><li>I have been playing the &quot;2048&quot; game for 5 years now. It is the only game I can stand on my phone.</li><li>I have stage fright. I am just really scared of my voice being the only thing that can be heard.</li><li>I have a long hair according to my hair dresser and my siblings.</li><li>When I was 12, I travelled to United Kingdom (UK) for an excursion. I had good and bad experiences and I can’t even remember most of them.</li><li>I love gadgets a lot.</li><li>I believe life is not interesting.</li><li>My dream is to live a simple life. Just work in a company with a good salary, wear heels, come back from work, eat, watch movies, sleep and donate my money.</li><li>When I was young, my favorite cartoon was &quot;Dora the explorer&quot;.</li><li>I was my brothers favorite goalkeeper. I was so good at it that my nickname was “Casillas”.</li><li>My favorite color is black. My wallpapers are black, my apps are in dark mode, my Bluetooth speaker is black.</li><li>I have social anxiety.</li><li>I dislike attention.</li><li>I can easily remember conversations from many years ago.</li><li>I am a hopeless romantic. I like watching romantic movies a lot, I like love songs and I believe everyone has a soul mate.</li><li>I am very bad at expressing myself. I tend to keep quiet so people won’t misunderstand me.</li><li>When i was 15, I learnt how to make clothes but I don’t really enjoy doing it.</li><li>I think interviewing people is fun.</li><li>I just decided one day that I am going to learn how to shoot an arrow, and play basketball and tennis.</li></ul><p>These facts basically summarizes my personality.</p><blockquote>Thank you for reading</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=71c4d721181a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>