<?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 Okwuosa Olisademe on Medium]]></title>
        <description><![CDATA[Stories by Okwuosa Olisademe on Medium]]></description>
        <link>https://medium.com/@okwuolisa322?source=rss-53091230bf50------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*avJwBFSR4yjbPwps</url>
            <title>Stories by Okwuosa Olisademe on Medium</title>
            <link>https://medium.com/@okwuolisa322?source=rss-53091230bf50------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 10:09:27 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@okwuolisa322/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[Let Me Give You the Tea… or Coffee?]]></title>
            <link>https://medium.com/@okwuolisa322/let-me-give-you-the-tea-or-coffee-bc6a81b963af?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc6a81b963af</guid>
            <category><![CDATA[matplotlib]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[seaborn]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Sat, 31 Oct 2026 21:33:16 GMT</pubDate>
            <atom:updated>2026-10-31T21:41:27.709Z</atom:updated>
            <content:encoded><![CDATA[<h3>Let Me Give You the Tea… or Coffee? — How Data Analysis Turns Raw Data Into Something You Can Actually Understand</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/949/1*og4TbuMHbJt91OOMgToWvw.jpeg" /></figure><p>There’s a lot one can do with the “tea” — or, in this case, coffee. Sometimes, data looks like a collection of numbers, dates and country names that don’t mean much on their own. But when you start grouping those numbers, comparing them and presenting them visually, a story begins to emerge.</p><p>This week, our focus was on <strong>grouping and aggregation</strong> with Pandas, alongside <strong>data visualization</strong> using Matplotlib and Seaborn. We explored how to organise data into meaningful categories, calculate useful summaries and turn the results into visual insights.</p><p>And what better way to put those skills to use than by investigating the international trade activities of a country?</p><h3>Getting to know the data: UN Comtrade</h3><p>One of the interesting resources we encountered was the United Nations International Trade Statistics Database, commonly known as UN Comtrade.</p><p>Website: <a href="https://comtradeplus.un.org/?utm_source=chatgpt.com">https://comtradeplus.un.org/</a></p><p>UN Comtrade provides international merchandise trade statistics reported by countries around the world. It allows users to explore imports and exports by reporter country, partner country, commodity, trade flow and period.</p><p>In simpler terms, you can investigate what a country imports, where those goods come from, what it exports, and the countries it trades with.</p><p>The teaching example for this week focused on the United Kingdom’s milk trade. We worked with data showing the UK’s milk imports and exports across different countries and periods. This provided a practical way to understand how grouping, aggregation and visualization work together.</p><h3>Grouping: putting similar data together</h3><p>Imagine having a spreadsheet containing thousands of records of milk imports and exports. Each row represents a trade record, with information such as the partner country, commodity code, trade flow, period and trade value.</p><p>Trying to understand the entire dataset by looking at individual rows would be tedious. This is where grouping becomes useful.</p><p>Grouping is the process of organising rows that share a common characteristic into separate groups. For example, we can group the milk data by trade flow to separate imports from exports, or by partner country to examine the UK’s trade with individual countries.</p><p>Pandas makes this possible with the groupby() method:</p><pre>grouped = milk.groupby(&#39;Trade Flow&#39;)</pre><p>This groups the records according to their trade flow. We can then access a particular group:</p><pre>grouped.get_group(&#39;Imports&#39;)</pre><p>This returns only the records relating to imports.</p><p>We can also group by more than one column:</p><pre>grouped = milk.groupby([&#39;Commodity Code&#39;, &#39;Trade Flow&#39;])</pre><p>Here, the data is organised according to combinations of commodity code and trade flow. This makes it possible to examine, for example, imports of one milk type separately from exports of that same milk type.</p><p>The important thing I learnt is that groupby() does not automatically calculate anything. It simply organises the data into groups so that we can perform further operations on them.</p><h3>Aggregation: getting useful information from the groups</h3><p>Once data has been grouped, the next step is often to calculate something meaningful from each group. This is called aggregation.</p><p>Aggregation involves applying summary calculations such as sum, average, minimum, maximum or count to grouped data.</p><p>For example, suppose we want to know the total value of milk imports from each partner country. We can group by partner and calculate the sum of the trade values:</p><pre>imports = milk[milk[&#39;Trade Flow&#39;] == &#39;Imports&#39;]<br><br><br>totalImports = (<br>    imports.groupby(&#39;Partner&#39;)[&#39;Trade Value (US$)&#39;]    <br>    .sum()    <br>    .sort_values(ascending=False)<br>)</pre><p>This code does three things:</p><ol><li>Filters the dataset to retain only imports.</li><li>Groups the remaining records by partner country and calculates the total trade value for each country.</li><li>Sorts the results from the highest trade value to the lowest.</li></ol><p>With this, we can identify the countries exporting the highest value of milk to the UK.</p><p>We also explored other aggregation functions, including mean(), min() and max(). These are useful when the question is not simply about total trade value but about average activity, the smallest transaction or the largest recorded value.</p><p>Another important concept was the Split–Apply–Combine approach: split the data into groups, apply a calculation or operation to each group, and combine the results into a useful output.</p><p>This is essentially the thinking behind much of the work we did with groupby(), aggregate() and apply().</p><h3>A little SQL along the way</h3><p>We also put our SQL knowledge to work through five practical exercises using the UK milk trade database. We used conditional statements to classify milk types, conditional aggregation to compare monthly imports and exports, and GROUPING SETS to produce quarterly reports with yearly subtotals. We also explored window functions to rank trading partners and calculate their percentage contributions, as well as Common Table Expressions (CTEs) to prepare a tidy dataset for a stacked bar chart.</p><p>These exercises helped me appreciate how SQL and Pandas can complement each other in data analysis. While Pandas allowed us to manipulate data in Python, SQL gave us the tools to retrieve, group, summarise and reshape information directly from a database.</p><h3>Taking the lesson home: Nigeria’s coffee trade in 2023</h3><p>The task was to recreate what we had learnt using another country of our choice. We were to select a country, choose a product or commodity, obtain its trade data for 2023 and investigate its dealings with other countries.</p><p>I chose my beloved country, Nigeria, and decided to investigate our coffee trade.</p><p>I selected coffee and a range of coffee-related products, using Nigeria as the reporting country, all partner countries, monthly data and both imports and exports for 2023.</p><p>This was where the exercise became particularly interesting. Rather than simply downloading a dataset and looking at the figures, I had to make decisions about which information was relevant and how best to organise it for analysis.</p><p>After downloading the data, I loaded it into Pandas and selected the columns I needed, including the reporting period, trade flow, partner country, commodity code, commodity description and trade value.</p><p>I also excluded the records where the partner was listed as World. This was important because the World figure represents an aggregate of trade with individual partners. Including it alongside the individual countries would have resulted in double-counting when calculating totals.</p><p>One interesting finding was that Indonesia was Nigeria’s largest import partner for the broad coffee category, with a trade value of approximately US$2.34 million in 2023.</p><p>Our exports, on the other hand, were considerably smaller. Egypt was the leading export partner for the broad coffee category, with a trade value of approximately US$77,104. The export records were also limited compared with imports, with only two partner countries appearing in the dataset.</p><p>Looking beyond the broad coffee category, I also examined the more specific commodity codes. The leading partners differed across the product categories, with countries such as China, Lebanon and Italy appearing as leading partners for some of them.</p><p>These findings helped me see how a single product category can contain several different commodities, each with its own trade pattern.</p><h3>Pivot tables: another way of looking at data</h3><p>Another important technique I worked with was the pivot table.</p><p>Before this week, my understanding of pivot tables was largely based on Microsoft Excel. Pandas’ pivot_table() initially felt different, but the underlying idea was familiar: reorganising data to make comparisons easier.</p><p>For example, I wanted to compare Nigeria’s coffee imports and exports across partner countries. I used:</p><pre>partner_summary = trade_countries.pivot_table(    <br>      index=&#39;partnerDesc&#39;,    <br>      columns=&#39;flowDesc&#39;,    <br>      values=&#39;primaryValue&#39;,    <br>      aggfunc=&#39;sum&#39;<br>)</pre><p>Here is what each argument does:</p><ul><li>index = &#39;partnerDesc&#39; places partner countries along the rows.</li><li>columns = &#39;flowDesc&#39; separates imports and exports into different columns.</li><li>values = &#39;primaryValue&#39; specifies the trade value to be calculated.</li><li>aggfunc = &#39;sum&#39; tells Pandas to add up the trade values for each combination.</li></ul><p>The result was a table with partner countries as rows and imports and exports as columns.</p><p>I then created a total trade column:</p><pre>partner_summary[&#39;Total Trade&#39;] = (    <br>        partner_summary[&#39;Import&#39;] +    <br>        partner_summary[&#39;Export&#39;]<br>)</pre><p>Finally, I sorted the table to identify the countries with the highest total trade values.</p><p>This was useful because instead of examining individual transactions, I could see the bigger picture of Nigeria’s coffee trade with each partner.</p><p>I also encountered missing values in some import and export combinations. For my analysis, I treated these missing values as zero, based on the assumption that a missing value represented no recorded trade for that particular flow. I recognised that this was an analytical assumption rather than something explicitly confirmed by the dataset, so it was important to document it.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xU76mldpoJuJ76CbV0dfxw.jpeg" /><figcaption>Nigeria’s coffee trade pivot table — 2023</figcaption></figure><h3>Making the findings visual</h3><p>Numbers in a table are useful, but visualisations can make comparisons and patterns much easier to recognise.</p><p>This week, I worked with Matplotlib and Seaborn to create charts. I learnt how to use bar charts to compare categories and line charts to examine changes over time.</p><p>For the UK milk dataset, I explored visualisations such as:</p><ul><li>Total imports compared with total exports.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/509/1*sUtgyksrZNywKPx0LUPtYQ.jpeg" /></figure><ul><li>The top five countries exporting milk to the UK.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/621/1*bbV_FST9NHFKqLOF0vcU6w.jpeg" /></figure><ul><li>The top five countries importing milk from the UK.</li><li>Processed versus unprocessed milk trade.</li><li>Monthly changes in milk imports and exports. etc</li></ul><p>For my Nigerian coffee analysis, I also plotted the leading import and export partners and examined monthly trade values.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/445/1*IKqLHwrjwNf3Bt_7mFLPOQ.jpeg" /></figure><h3>Challenges, lessons and takeaways</h3><p>Pivot tables were probably my biggest initial challenge. They were different from the Excel approach I was accustomed to, and understanding how index, columns, values and aggfunc worked together took some time.</p><p>Reading the Pandas documentation, experimenting with different arguments and examining the resulting DataFrames helped me understand the method better.</p><p>Another challenge was deciding what to do with missing values. It is tempting to replace every missing value with zero, but missing data does not always mean zero. Understanding the difference and documenting the assumption was an important part of the exercise.</p><p>I also learnt to pay attention to the structure of a dataset before analysing it. Excluding the World aggregate, selecting relevant columns and checking the available commodity codes made the subsequent analysis more reliable.</p><p>Perhaps my biggest takeaway from this week is that <strong>data analysis is not just about knowing which function to use; it is about knowing which question you want the data to answer.</strong></p><p>Grouping helps us organise information. Aggregation helps us summarise it. Pivot tables help us reshape it for comparison. Matplotlib and Seaborn help us communicate what we have discovered.</p><p>And sometimes, a simple question such as “Who does Nigeria buy most of its coffee from?” can lead to a much more interesting investigation.</p><p>There is still a lot more to explore, but this week has given me a better understanding of how raw trade records can be transformed into meaningful insights.</p><p>After all, there’s a lot one can do with the tea. And apparently, quite a lot with coffee too!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc6a81b963af" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[GDP, Weather and a Few Headaches: My Week 7]]></title>
            <link>https://medium.com/@okwuolisa322/gdp-weather-and-a-few-headaches-my-week-7-e27b7981b0aa?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/e27b7981b0aa</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Sat, 24 Oct 2026 19:02:02 GMT</pubDate>
            <atom:updated>2026-10-24T19:26:52.836Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 7 was mainly about taking data that was not immediately ready to use, cleaning it up, bringing different pieces of information together, and then trying to make sense of what the data was telling me.</p><p>This week was also where I started seeing how some of the things I had been learning in SQL could be done in pandas, sometimes in a completely different way.</p><h3>What I learnt</h3><p>One of the biggest things I learnt this week was how to work with dates in pandas.</p><p>I learnt that a date can be made the index of a DataFrame, which makes it much easier to work with a particular period. For example, I was able to select the weather records between 1 April and 11 April 2014 and also look at the weather during the spring months of March, April and May.</p><p>I also worked with missing data. Instead of just ignoring missing values, I learnt how to find them and decide what to do with them. For example, I checked the weather data for missing values in <strong>Max Gust SpeedKm/h</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LiXeLdnobloTaHT-RvHF_w.jpeg" /><figcaption>The notebook output showing the rows with missing Max Gust SpeedKm/h</figcaption></figure><p>Another thing I found interesting was discovering that pandas has functions that can do things similar to things I had previously done in SQL.</p><p>For example:</p><ul><li>nlargest() can be used to find the highest values, similar to using TOP with an ORDER BY in SQL.</li><li>nsmallest() does the opposite.</li><li>isin() can be used when checking whether values belong to a particular list, similar to IN in SQL.</li><li>isna() helps find missing values.</li><li>dropna() removes rows with missing values.</li><li>merge() allows different datasets to be brought together using a common column.</li><li>sort_values() can be used to arrange data in a particular order.</li><li>loc[] makes it possible to select specific rows or ranges of data.</li></ul><p>I liked seeing this because it helped me realise that although SQL and pandas look different, they can often be used to solve similar problems.</p><h3>Looking at the weather data</h3><p>I also worked with London’s weather data for 2014.</p><p>I looked at the maximum, mean and minimum temperatures throughout the year. Plotting them made it easier to see how the temperatures changed rather than just looking at rows of numbers.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-QfWHZnLtziGTPzK79bXpw.jpeg" /><figcaption>The London temperature plot showing Max TemperatureC, Mean TemperatureC and Min TemperatureC</figcaption></figure><p>I also looked specifically at the mean humidity during spring.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qfuKOWf09Dap4TpWR0VF0Q.jpeg" /><figcaption>The spring Mean Humidity plot from March, April and May</figcaption></figure><p>These exercises helped me understand why visualising data can be useful. A table can contain a lot of information, but a plot can make a pattern much easier to notice.</p><h3>Combining information about different countries</h3><p>Another part of the week involved bringing information about different countries together.</p><p>I worked with GDP and life expectancy data and combined them using the country name. This allowed me to look at the two pieces of information together instead of treating them as completely separate datasets.</p><p>I then looked at whether countries with higher GDP generally also had higher life expectancy.</p><p>The scatter plot made the relationship easier to see. There was a general upward pattern, but it was not perfect. Some countries did not follow the general pattern.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D1_Gt3UwTwomcNmMM5bvNw.jpeg" /><figcaption>The GDP (£m) vs Life expectancy (years) scatter plot</figcaption></figure><p>I also used a statistical test to measure the relationship between GDP and life expectancy. The result showed a positive relationship between the two.</p><p>What I found interesting here was that the numbers gave me a way to measure something that I had already started noticing from the plot.</p><h3>Going a step further</h3><p>I also compared countries at the top of the GDP ranking with countries at the top of the life expectancy ranking to see how much they overlapped.</p><p>I then looked at countries that had relatively high GDP but life expectancy below 60 years. Nigeria and South Africa stood out from the data.</p><p>That made me think about something that is easy to overlook when looking at economic figures: a country’s overall wealth does not necessarily tell the full story about the quality of life of the people living there.</p><p>There can be differences in healthcare, access to services, living conditions and how wealth is distributed. So even when two countries have significant economic activity, the experience of the people living there can be very different.</p><h3>GDP per person</h3><p>I also took the analysis a step further by looking at GDP per person rather than just the total GDP of a country.</p><p>This made more sense for some comparisons because countries have very different population sizes. A country can have a very large economy simply because it has a very large population.</p><p>While doing this, I ran into an interesting problem.</p><p>I had rounded population figures to millions earlier in the analysis. For very small countries, the rounded value became 0.0. When I tried to use that rounded figure to calculate GDP per person, it resulted in an inf value.</p><p>That forced me to go back and use the actual population figures instead of the rounded ones.</p><p>It was a small mistake, but it was a useful lesson: <strong>a number that looks fine for displaying results is not necessarily suitable for doing further calculations.</strong></p><h3>My Week 7 assessment</h3><p>For the assessment, I chose to look at whether there was a relationship between a country’s GDP per capita and the percentage of its population using the Internet.</p><p>I used data for 2023, cleaned the two datasets, removed the regional and world aggregates, and then brought the country data together.</p><p>The final dataset contained 182 countries.</p><p>The result showed a strong positive relationship between GDP per capita and Internet usage. In simple terms, countries with higher GDP per person generally tended to have a higher percentage of their population using the Internet.</p><p>The result was statistically significant, but I also learnt that a relationship between two things does not automatically mean that one directly causes the other.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*u-x4Lk-o78NkrDzRIlRO4g.jpeg" /><figcaption>Your GDP per Capita vs Internet Usage in 2023 scatter plot</figcaption></figure><p>I think this was one of the more interesting parts of the week because the question was simple enough to understand without knowing the technical details, but the analysis showed how data can be used to investigate questions like it.</p><h3>Challenges</h3><p>There were quite a few moments where I got stuck this week.</p><p>Some of the problems were simple mistakes, like trying to treat GMT as a normal column when it was actually the index of the DataFrame.</p><p>There were also problems with the World Bank datasets. The downloaded files contained information before the actual data, so they could not simply be loaded with the usual read_csv() approach. I had to work out how to load the actual data and then remove the regional and world aggregates before doing the analysis.</p><p>I also had to deal with different datasets having different numbers of countries. Getting them into a form where they could be compared required some cleaning and merging.</p><p>And then there was the GDP per person problem where rounding the population caused some countries to have a population of 0.0 million.</p><h3>How I overcame them</h3><p>The biggest thing that helped was stopping and looking at what the data actually contained instead of assuming that it was structured the way I expected.</p><p>I checked column names, indexes, shapes and outputs whenever something did not work.</p><p>I also went back to earlier steps when I realised that a later calculation depended on information I had already changed.</p><h3>One thing I am taking away from Week 7</h3><p>If I am being honest, I am also rueing the fact that I waited until close to the deadline to properly get through some of this work.</p><p>Trying to finish assignments, understand new concepts, fix errors and complete the assessment around the same time exposed me to something I probably needed to experience.</p><p>Knowing how to write the code is one thing. Being able to sit with an unfinished task, find where things went wrong, understand the data, fix the problem and eventually get to a result is another.</p><p>Waiting until the deadline made that much harder than it needed to be.</p><p>So beyond the pandas and data analysis itself, Week 7 has made one thing very clear to me: <strong>leaving unfinished work until the last minute doesn’t just create pressure. It can also make the actual learning harder.</strong></p><p>I would rather deal with the work gradually, leave room for mistakes, and have enough time to actually understand what I am doing.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e27b7981b0aa" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 6 at DataraFlow: From IP Addresses to DataFrames]]></title>
            <link>https://medium.com/@okwuolisa322/week-6-at-dataraflow-from-ip-addresses-to-dataframes-b1ecf0254a28?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/b1ecf0254a28</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Mon, 19 Oct 2026 10:15:08 GMT</pubDate>
            <atom:updated>2026-10-19T10:15:08.175Z</atom:updated>
            <content:encoded><![CDATA[<p>This week was <strong>intense chaos from start to finish. </strong>It was<strong> </strong>easily the most intense week of the program so far.</p><p>From dynamic IP issues, to the largest coursework we have had, to a new quiz format, there was always something to figure out.</p><p>But I conquered them all this week.</p><p>And strangely enough, one of my biggest lessons started before I even got to SQL, NumPy or Pandas.</p><h3>The Week Started With an IP Address</h3><p>Before this week, I didn’t know much about IP addresses.</p><p>That changed when we had to connect to a live SQL Server.</p><p>For security reasons, our IP addresses needed to be <strong>whitelisted</strong>, meaning the server had to recognise and allow our addresses before we could connect.</p><p>My problem was that I had a <strong>dynamic public IP</strong>, which could change.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hcu9ZYZ2ErljJRdeMORfHA.jpeg" /><figcaption>The error that started my IP-address adventure.</figcaption></figure><p>That sent me down a completely unexpected rabbit hole.</p><p>I learnt the difference between <strong>public and private IP addresses</strong> and started looking for ways to make my dynamic public IP effectively static. Most of the solutions I found involved premium VPN subscriptions, so I started looking for alternatives.</p><p>People helped me a lot here.</p><p>My mentor in the cohort suggested <strong>Cloudflare</strong>, while my brother suggested creating a virtual machine with Oracle.</p><p>I was actually in the process of creating the virtual machine when I received an email from the DataraFlow team saying that a number of IP addresses had been whitelisted across different <strong>ISPs (Internet Service Providers)</strong> and that we should try connecting again.</p><p>Mine was among them.</p><p>Huge relief.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*V_X3L3jOFW0HGoYflojVpA.jpeg" /><figcaption>Successful database connection</figcaption></figure><p>I didn’t plan to learn about networking this week, but connecting to a live server made sure I did.</p><h3>Finally, SQL</h3><p>Once I could connect, I could finally focus on what I was actually there for.</p><p>SQL this week went beyond simply writing SELECT * FROM table.</p><p>We used <strong>Microsoft SQL Server</strong>, with our datasets stored under the week06 schema. I learnt how to inspect the database using INFORMATION_SCHEMA instead of blindly guessing table and column names.</p><p>This became useful because some of the weather columns contained spaces in their names and needed square brackets:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*578XGTjdMEwovAZjgCFrWg.jpeg" /></figure><p>We worked with filtering, sorting and grouping data using things like WHERE, ORDER BY, GROUP BY and HAVING.</p><p>We also covered handling missing values with NULL, COALESCE() and NULLIF(), using CASE for conditional results, converting data types, joins, CTEs, window functions and PIVOT.</p><p>It was a lot to take in, especially as the queries became longer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*LjuJLMy2MbwKAfYpmzAQdg.jpeg" /></figure><p>One task that stood out involved weather data from six cities: London, Beijing, Brasilia, Cape Town, Delhi and Moscow.</p><p>Each city had its own table containing 365 rows. We needed to select the city, date and mean temperature and combine everything into one dataset.</p><p>This introduced a simple but important idea:</p><pre>SELECT<br>    &#39;London&#39; AS city,<br>    GMT,<br>    [Mean TemperatureC]<br>FROM week06.london_2014</pre><p>There was no city column containing London in the original table. &#39;London&#39; AS city created that label in the result.</p><p>Using UNION ALL, we could do the same for all six cities and stack the results.</p><p>Six cities × 365 days gave us <strong>2,190 rows</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p5CrqltfEhnX3lB1ynJjAg.jpeg" /><figcaption>UNION ALL query</figcaption></figure><p>This was also where SQL started connecting with Python. Instead of running a query and stopping there, we could retrieve the result directly into a Pandas DataFrame.</p><p>That connection between the tools would become one of my biggest lessons from the week.</p><h3>NumPy: More Than Just Arrays</h3><p>Next came <strong>NumPy</strong>.</p><p>We started with arrays and their dimensions — 0D, 1D, 2D and 3D — before moving into indexing and slicing.</p><p>The deeper we went, the more there was to understand.</p><p>I worked with data types using dtype, converting arrays using astype(), and learnt the difference between a <strong>copy and a view</strong>.</p><p>Then came reshaping.</p><pre>newarr = arr.reshape(2, 2, -1)</pre><p>Using -1 tells NumPy to calculate that dimension for me. The catch is that I can only have one unknown dimension.</p><p>We also covered joining arrays with concatenate(), stack(), hstack(), vstack() and dstack(), before moving into splitting, sorting and searching.</p><p>One function that initially needed a little more attention was searchsorted().</p><p>Rather than simply checking whether something exists, it finds where a value should be inserted into a sorted array while maintaining the order.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/588/1*sYoxDz1FQWAxWKroXO8nVg.jpeg" /><figcaption>NumPy exercise using searchsorted()</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1PKEVlAWlNEIYmvOCQ1b4g.jpeg" /><figcaption>Some numpy tasks</figcaption></figure><p>We also worked with NumPy arithmetic and statistical functions for things like minimums, maximums, averages, medians and percentiles.</p><p>The new quizzes helped expose some gaps too.</p><p>There were moments where I thought I understood a concept until a question presented it differently. Going back to understand <em>why</em> I got something wrong was more useful than simply memorising the correct answer.</p><h3>Then Came Pandas</h3><p>After NumPy came <strong>Pandas</strong>, and things started connecting even more.</p><p>We queried WHO population and tuberculosis data directly from the database:</p><pre>df = run_query(&quot;SELECT * FROM week06.who_pop_tb_some&quot;)</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jZtqHjyMXdBjVwfvCJjTzA.jpeg" /></figure><p>The result came into Python as a DataFrame, where we could continue analysing it.</p><p>I worked with selecting rows and columns, iloc, head(), tail(), sorting data and performing calculations such as sum, minimum, maximum, mean and median.</p><p>We also worked with filtering.</p><p>For example:</p><pre>df[df[&#39;Population (1000s)&#39;] &gt; 80000]</pre><p>Instead of manually going through the data, I could simply state the condition I wanted.</p><p>Multiple conditions could also be combined using &amp; for AND and | for OR.</p><p>We also calculated values using existing columns, including TB deaths per 100,000 people.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BhgrP-06nYiKQy-21vaJlw.jpeg" /><figcaption>some pandas tasks</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ktt86oh0Fu9A5z6fLrBk2w.jpeg" /><figcaption>other tasks involving pandas</figcaption></figure><p>This was where the relationship between everything became much clearer to me:</p><p><strong>SQL could retrieve the data. Pandas could organise, filter and analyse it. NumPy provided many of the numerical tools behind that work.</strong></p><p>They were no longer three completely separate things I was learning.</p><h3>Challenges and How I Handled Them</h3><p>My biggest challenge this week was the <strong>volume of work</strong>.</p><p>This was the largest coursework we have had so far, and it came with server connection problems and a new quiz format.</p><p>The IP problem couldn’t be solved by correcting a line of code. I had to research, ask questions and explore solutions I had never used before.</p><p>For the actual coursework, breaking things down helped.</p><p>When multidimensional indexing became confusing, for example, I stopped trying to memorise the output and started looking at what each index was actually selecting.</p><p>The same applied to SQL.</p><p>When a long query looked confusing, understanding each section individually made the whole thing less intimidating.</p><p>One thing I would do differently is spend less time trying to remember syntax and more time understanding what each operation is trying to achieve. Syntax can always be checked. Understanding what you want the computer to do is the more important part.</p><h3>Week 6: Done</h3><p>This week had everything.</p><p>Dynamic IP problems. A live SQL Server. SQL queries. NumPy. Pandas. The largest coursework so far. A new quiz format. And enough errors to keep things interesting.</p><p>But I got through it.</p><p>My favourite part was probably seeing the tools finally begin to connect. Instead of learning SQL today, NumPy tomorrow and Pandas the next day as completely separate topics, I could see how they fit into an actual workflow.</p><p>And I learnt something else too.</p><p>Sometimes, the thing stopping you from analysing your data isn’t your Python code.</p><p>Sometimes it’s your IP address. 😂 😂 😂</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wZ9S41uRrckKadv6kQSFvQ.jpeg" /><figcaption>Favourite result from Week 6</figcaption></figure><p>Week 6 done.</p><p>Definitely not peacefully, but done.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b1ecf0254a28" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Thing I Hoped I’d Never Have to Do Again]]></title>
            <link>https://medium.com/@okwuolisa322/the-thing-i-hoped-id-never-have-to-do-again-ca9cfd2731a3?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/ca9cfd2731a3</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Sun, 11 Oct 2026 08:13:39 GMT</pubDate>
            <atom:updated>2026-10-11T08:13:39.679Z</atom:updated>
            <content:encoded><![CDATA[<p>My undergraduate research and project writing created this phobia in me that after I finished, I dreaded anything that had to do with research writing. In fact, it made me reluctant to further my studies because I knew I would definitely meet it somewhere ahead.</p><p>So, you can guess how shocked I was when I saw research writing in the curriculum for this internship.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/228/1*P1ZmPpPKw9wnvZkwRBk8LQ.jpeg" /></figure><p>Each week in DataraFlow usually has something different for us to learn and write about, but Weeks 4 and 5 were different. The two weeks were combined and introduced us to research writing. The goal was to lay the groundwork for choosing our individual research topics and eventually submitting a proposal (solving a problem using machine learning).</p><p>It was an insane period for me.</p><p>In fact, I barely did anything during the first week.</p><h3>Week 4: Avoidance</h3><p>I wish I could say I immediately faced the challenge and got to work, but that wasn’t what happened.</p><p>Week 4 came, and that old fear of research writing came with it. I found myself putting off the work. The more I thought about everything involved, the less I wanted to start.</p><p>And that was basically how most of Week 4 went.</p><p>As we entered Week 5, though, I knew this ought not to be so. Avoiding it wasn’t going to make it disappear. If I wanted to complete the internship, I had to do the work.</p><p>More importantly, I knew I had to conquer this challenge and face my fear head-on.</p><p>And boy, did I do that!</p><h3>Starting With What I Knew</h3><p>One thing that made the process easier was starting from a niche I was already familiar with: <strong>waste management</strong>.</p><p>For context, I studied Environmental Technology at the Federal University of Technology, Owerri (FUTO), so instead of jumping into a completely unfamiliar area, I decided to play in an environment I already knew.</p><p>That gave me somewhere to start.</p><p>But knowing the field didn’t mean I automatically had a research topic. I still had to figure out what exactly I wanted to study, what other people had already done, what was missing, and whether I could actually build something around it.</p><p>That was where the real work began.</p><h3>Down the Research Rabbit Hole</h3><p>I started looking for research papers around waste management.</p><p>One paper led to another. Then another.</p><p>I had to read what previous researchers had done, understand the problems they were trying to solve, look at the methods they used and, most importantly, ask myself:</p><p><strong>What is still missing?</strong></p><p>This part took time.</p><p>There were hours of researching, studying, questioning and brain-racking. Sometimes I thought I had found something useful, only to realise I needed to dig deeper. Other times, something I read would lead me in an entirely different direction.</p><p>Slowly, though, things started coming together.</p><h3>Finding My Research Gap</h3><p>One of the biggest things I learned during these weeks was that choosing a research topic isn’t just about finding something that sounds interesting.</p><p>There has to be a reason for doing the research.</p><p>So rather than simply saying, <em>“I want to use machine learning for waste management,”</em> I had to look at what had already been done and find something worth investigating further.</p><p>After going through several studies and comparing their work, I eventually identified a gap I could work with.</p><p>That was a major turning point for me.</p><p>Suddenly, I wasn’t just reading random papers anymore. I had a direction.</p><h3>From an Idea to a Proposal</h3><p>Finding the gap was only part of the job.</p><p>I still had to turn everything I had discovered into a research topic and then explain what I intended to do in a proposal.</p><p>That meant making sure the problem, previous research, available data and the machine learning side of the work actually connected.</p><p>I also had to be realistic. This is still a learning process, so whatever I proposed had to be something I could actually work on with the resources and knowledge available to me.</p><p>After all the searching, reading, comparing, questioning and rewriting, I finally came up with my topic.</p><p>Then I submitted it for approval.</p><p>And waited.</p><h3>The Approval Mail</h3><p>A few hours later, I received the approval mail.</p><p>I can’t explain the amount of joy that flooded my soul when I saw it. It was a surreal moment.</p><p>Of course, getting the topic approved doesn’t mean the research is finished. Far from it. There is still plenty of work ahead.</p><p>But for me, that email meant something more.</p><p>A week earlier, I was barely touching the work because of how much I dreaded research writing. Now, I had gone through research papers, found a problem I wanted to investigate, identified a gap, developed a topic and written a proposal that got approved.</p><p>That felt good.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/204/1*huoSMH2Jt5KLZyLhQZOiCA.jpeg" /></figure><p>Really good.</p><h3>What Weeks 4 &amp; 5 Taught Me</h3><p>These two weeks were less about Python code and more about learning how to ask the right questions.</p><p>What has already been done?</p><p>What is missing?</p><p>Is there data I can use?</p><p>What exactly am I trying to find out?</p><p>And can I turn all of that into something I can actually work on?</p><p>I also learned something about myself.</p><p>The fear I carried from my undergraduate project was still there, but avoiding research only made it look bigger than it actually was. Once I started breaking the work into smaller pieces and dealing with them one after another, it became much more manageable.</p><p>I won’t suddenly pretend that research writing is now my favourite thing in the world.</p><p>But I don’t dread it the way I did before Week 4.</p><p>And I’ll take that as a win.</p><h3>What Comes Next</h3><p>The proposal is approved, but that’s only the beginning.</p><p>Next comes the actual research: working with the data, building and testing the machine learning models, analysing what I find and eventually putting everything together into a proper research paper.</p><p>There is still a long way to go.</p><p>But considering how Week 4 started, I’m looking forward to seeing where this goes.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ca9cfd2731a3" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 3: Another Week, Another Layer Unpacked]]></title>
            <link>https://medium.com/@okwuolisa322/week-3-another-week-another-layer-unpacked-e468d94e30a8?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/e468d94e30a8</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Thu, 24 Sep 2026 03:59:26 GMT</pubDate>
            <atom:updated>2026-09-24T03:59:26.779Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/156/1*IeuBKZToQR_mpEpLrnXInw.jpeg" /></figure><p>This was my mood for most of the week: <em>“Is that all you’ve got?”</em></p><p>I came into the week enthusiastic, highly anticipating the challenges that were going to come my way. I was ready for Python to throw some serious punches.</p><p>Well… <em>the punches turned out to be more like gentle taps on the shoulder.</em></p><p>Here’s what we covered.</p><h3>Modules — Python’s Toolbox</h3><p>Modules are simply Python files containing useful code that we can reuse.</p><p>Instead of writing everything ourselves, we can import a module and use what it already provides.</p><p>For example:</p><p>import math</p><p>where;</p><p><strong>import</strong> is a Python keyword used to bring a module into your program.</p><p><strong>math</strong> is a built-in Python module that provides mathematical functions and constants.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*sFmHeQpR5TFBvyZeIwk_fw.jpeg" /></figure><h3>Dates — Teaching Python About Time</h3><p>We worked with dates and times using the <strong>datetime</strong> module.</p><p>I learned how to get the current date, format dates, convert strings into dates, and calculate the difference between dates.</p><p>Things like:</p><p><strong>datetime.now()</strong> and <strong>strftime() </strong>started making much more sense.</p><p>Apparently, Python can tell me what day it is.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/219/1*PDWfVSMCxb0uwPUC_DLieg.jpeg" /></figure><h3>Maths — Let Python Do the Counting</h3><p>We worked with things like <strong>math.pi</strong>, rounding numbers using the <strong>round() </strong>function and other mathematical operations.</p><p>For example, instead of typing out the value of π every time:</p><p><strong>math.pi </strong>does the job.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/449/1*miJH2KLdGGbIPSKKtmVozg.jpeg" /></figure><h3>JSON — Getting Python to Speak Data</h3><p>I learned how to convert Python data into JSON (JavaScript Object Notation) and convert JSON back into Python data.</p><p>The main functions were:</p><p><strong>json.loads()</strong>: Converts a JSON string into a Python object<br><strong>json.dumps()</strong>: Converts a Python object into a JSON string</p><h3>pip, Packages, try/except and Virtual Environments</h3><p>I learned that <strong>packages</strong> are collections of reusable Python code, while <strong>pip</strong> helps us install and manage those packages.</p><p><strong>try/except </strong>taught me how to handle errors without allowing my program to completely lose its mind.</p><p>A<strong> virtual environment</strong> gives a project its own space for its packages and dependencies, helping prevent different projects from interfering with one another.</p><h3>Reading and Writing Files</h3><p>We also learned how Python can interact with files.</p><p>Python can open a file, read what’s inside it, write new information, and even add information without replacing what was already there.</p><p>Interesting right?</p><h3>And Then Came SQL</h3><p>I was particularly excited about SQL (Structured Query Language).</p><p>We had an introduction to SQL and databases, and I’m highly anticipating diving deeper into it.</p><p>SQL is a language I genuinely enjoy, so getting back into it within the context of this programme was something I looked forward to.</p><h3>A Little Responsibility on the Side</h3><p>Outside the technical work, I also volunteered to be the acting class captain for my group.</p><p>Taking on the role has made me more active and accountable.</p><p>I now have an extra reason to pay attention, communicate with others, follow up on things and make sure I’m playing my part.</p><p>It’s a small role, but I’m enjoying the responsibility and the opportunity to contribute to the group beyond my individual learning.</p><h3>Week 3, Done.</h3><p>Overall, it was a good week. I came in expecting Python to throw everything at me, but the week turned out to be more of a steady walk than a wrestling match.</p><p>I learned new concepts, got more comfortable working with Python, took on a little more responsibility within my group.</p><p>And honestly, I’m looking forward to going deeper into SQL.</p><p><strong>Week 3 done. Week 4, let’s go.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e468d94e30a8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 2 at DataraFlow: Finding My Rhythm Again]]></title>
            <link>https://medium.com/@okwuolisa322/week-2-at-dataraflow-finding-my-rhythm-again-142818d7a19c?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/142818d7a19c</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Sat, 19 Sep 2026 07:35:52 GMT</pubDate>
            <atom:updated>2026-09-19T07:35:52.808Z</atom:updated>
            <content:encoded><![CDATA[<p>Week 2 at DataraFlow was an interesting one. It was also an emotional week for me.</p><p>It started with my team, Manchester United, losing 1–0 in the Manchester derby. Then came another loss to Brighton, which saw us exit the Carabao Cup.</p><p>At that point, I wasn’t exactly starting the week with the best energy.</p><p>But gradually, I found my flow again. The highlight of the week was the Library Management system I built. Details about it down below</p><p><strong>Getting Deeper Into Python</strong></p><p>This week took me further into Python and programming.</p><p>Unlike some of the earlier concepts where I was mainly learning how to write instructions for a computer, I started learning how to organize those instructions into systems that resemble things in the real world.</p><p>The concepts sounded intimidating at first:</p><p><strong><em>OOP. Inheritance. Iterators. Polymorphism. Scope.</em></strong></p><p>But as the week progressed, they began to make more sense. Here’s how I understand them now.</p><p><strong>Object Oriented Programming: Thinking About Things</strong></p><p>One of the biggest ideas I learned this week was Object Oriented Programming, commonly called OOP.</p><p>A simple way I think about it is this:</p><p>Imagine you’re designing software for a library. Instead of thinking only about instructions, you start thinking about the things that exist in the library. There are:</p><p><em>books</em></p><p><em>members.</em></p><p><em>different types of books.</em></p><p><em>different types of members. etc</em></p><figure><img alt="An image describing a class and possible contents." src="https://cdn-images-1.medium.com/max/462/1*3-AOyjbQYhMWQt7BBO-MZA.jpeg" /></figure><p>Each of these things has information about it and things it can do.</p><p>OOP gives programmers a way to represent these real-world things inside a program.</p><p>It’s also about organizing the problem you’re trying to solve.</p><p><strong>Inheritance: Building on What Already Exists</strong></p><p>Then came inheritance. <em>I don’t know why Esau and Jacob came to my mind when I first saw inheritance, lol.</em></p><p>The easiest way I can explain it is by thinking about a family.</p><p>Imagine we have a general category called Vehicle.</p><p>A car is a <em>vehicle</em>.</p><p>A bicycle is also a <em>vehicle</em>.</p><p>They have things in common, but they aren’t exactly the same.</p><p>Inheritance allows us to create a general structure and then build more specific things from it.</p><p>This means we don’t have to start from scratch every time.</p><p>I found this concept particularly interesting because it reflects something we see everywhere in the real world: things can share characteristics while still being different.</p><figure><img alt="A parent class with its sub classes (child class)" src="https://cdn-images-1.medium.com/max/846/1*0MNEjp6Nc6NMU5rNeSu-OQ.jpeg" /></figure><p><strong>Iterators: Taking Things One Step at a Time</strong></p><p>Iterators were one of the concepts I had to spend more time with.</p><p>The basic idea is actually quite simple.</p><p>Imagine having a queue of 100 people.</p><p>Instead of dealing with everyone at once, you could take them one person at a time.</p><p>That’s roughly the way I began to understand iterators.</p><p>They allow a program to move through a collection of things, one item at a time.</p><p>This became clearer when I started working with concepts such as iter(), next() and knowing when there were no more items to process.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/590/1*km-yNv6nsVUoRXO34EBewA.jpeg" /></figure><p><strong>Polymorphism: Same Instruction, Different Behaviour</strong></p><p>Polymorphism???</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/239/1*oZFgZNvw7Gbwukd1KcmhDg.jpeg" /></figure><p>This initially sounded like one of those programming words designed to scare beginners.</p><p>But the idea became much simpler when I connected it to everyday life.</p><p>Imagine telling a car to “move.”</p><p>It drives.</p><p>Tell a bicycle to “move.”</p><p>It pedals.</p><p>The instruction is essentially the same, but the response depends on what you’re dealing with.</p><p>That’s the basic idea I took away from polymorphism:</p><p><strong>Different things can respond to the same instruction in different ways.</strong></p><p>This was one of those concepts where the name sounded more complicated than the actual idea.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/538/1*4GEZ5Z1WD2HWDUz4G1WnIw.jpeg" /></figure><p><strong>Scope: Understanding Where Things Belong</strong></p><p>The final major topic was scope.</p><p>This was about understanding where information or variables existwithin a program and where they can be accessed.</p><p>A real life comparison would be an office. There might be information available to the entire company, information available only to a particular department, and information available only to one person.</p><p>Programming has similar boundaries.</p><p>I learned about <strong>local, enclosing, global and built-in scope</strong>, as well as the global and nonlocal concepts.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/449/1*jo-vClnFjp8wNVJkRqIeQg.jpeg" /></figure><p><strong>The Project: Building a Library Management System</strong></p><p>The part of this week that brought everything together was the <strong><em>Library Management System project</em></strong><em>.</em></p><p>The task was essentially to think about what a real library would need and then design a program around it.</p><p>At the center of the system were books and members.</p><p>But even that wasn’t as simple as it first sounded.</p><p>There are different kinds of books.</p><p>There are electronic books and physical books.</p><p>There are different kinds of members.</p><p>There are students and teachers.</p><p>Each one has information that makes it different from the others.</p><p>The challenge was to design a system that could represent these differences while still keeping the common characteristics together.</p><p>The project wasn’t just about storing information.</p><p>The library also needed to be able to perform activities.</p><p>Books could be:</p><ul><li><em>Available</em></li><li><em>Borrowed</em></li><li><em>Searched for</em></li><li><em>Returned</em></li></ul><p>Members could:</p><ul><li><em>Borrow books</em></li><li><em>Have borrowing limits</em></li><li><em>Keep track of the books they currently have</em></li></ul><p>The system also needed to distinguish between different types of members and books.</p><p>This made me think less about individual lines of code and more about the design of the entire system.</p><p>What should belong together?</p><p>What should be shared?</p><p>What should be different?</p><p>What happens when one thing changes?</p><p>Those were some of the questions behind the project.</p><p><strong>Where the Week’s Concepts Came Together</strong></p><p>Looking back at the project, I could see the role each topic played.</p><p><strong><em>OOP</em></strong> helped me think about the different things that exist in the library.</p><p><strong><em>Inheritance</em></strong> helped me represent relationships between general categories and more specific ones.</p><p><strong><em>Polymorphism</em></strong> allowed different types of objects to behave differently while responding to the same kind of instruction.</p><p><strong><em>Iterators</em></strong> helped with moving through collections of books and borrowed books.</p><p><strong><em>Scope</em></strong> helped me understand how information is managedand accessed within the program.</p><p>This was probably the biggest lesson for me.</p><p>The topics weren’t really separate.</p><p>They were<em> pieces of one larger system</em><strong>.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/975/1*AqF5njuak9BU9kzKhUR-bQ.jpeg" /><figcaption>code snippet showing my project structure</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/809/1*9HsGWwWQZr5AYDg0fjAFQA.jpeg" /><figcaption>Another one</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/708/1*RRltpFQ7ALqKIdeC5HmHiA.jpeg" /><figcaption>Screenshot showing the project running</figcaption></figure><p><strong>When You Don’t Understand, Find Another Explanation</strong></p><p>One thing that helped me significantly this week was realizing that sometimes the problem isn’t that you can’t learn something; it’s that the explanation you’re currently using hasn’t clicked yet.</p><p>There were concepts I went through several times before they made sense.</p><p>Shoutout to my colleagues that provided materials that were extremely he;pful to me; making me understand things better and regain some momentum.</p><p><strong>The video that helped me:</strong></p><p>“Python Classes Explained (2+ Hour Full Course)”<br> <strong>Creator:</strong> Python Made Easy</p><p>👉 <strong>Video: </strong><a href="https://youtu.be/LAiNEkggiU4?si=ia3JVYGzGCIf9S-N">https://youtu.be/LAiNEkggiU4?si=ia3JVYGzGCIf9S-N</a></p><p><strong>What I Took Away From Week 2</strong></p><p>If I had to summarize this week in one sentence, it would be:</p><p><em>I moved from simply learning Python concepts to beginning to understand how those concepts can be used to build something</em><strong>.</strong></p><p>I also learned that getting stuck isn’t necessarily a sign that I’m not making progress.</p><p>My Club’s results may not have improved my mood at the beginning of the week, but eventually, I got back into my learning rhythm.</p><p>And that’s something I’m taking with me into the next week.</p><p>I’m still very much a learner.</p><p>There are concepts I understand today that I couldn’t explain a few weeks ago, and there are concepts I’m still trying to fully grasp.</p><p>But that’s the point of the journey.</p><p>By the way, i am now active in my group. Promised myself I’ll do that in week one and it’s been good.</p><p><strong>Image Credits</strong><br> Some of the illustrations used in this article were sourced from the following creators:<br> • Pawan Natekar<br> • Subhadra Bhupathiraju<br> • Rachit Tayal<br> • GeekPython</p><p>Thank you to the original creators for making these resources available.</p><p><strong>Week 2: Done.</strong></p><p><strong>On to Week 3. 🚀</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=142818d7a19c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Week 1 at DataraFlow: A Promising Start to an Exciting Journey]]></title>
            <link>https://medium.com/@okwuolisa322/week-1-at-dataraflow-a-promising-start-to-an-exciting-journey-2932a3197940?source=rss-53091230bf50------2</link>
            <guid isPermaLink="false">https://medium.com/p/2932a3197940</guid>
            <dc:creator><![CDATA[Okwuosa Olisademe]]></dc:creator>
            <pubDate>Sun, 06 Sep 2026 21:57:36 GMT</pubDate>
            <atom:updated>2026-09-06T21:57:36.124Z</atom:updated>
            <content:encoded><![CDATA[<p>Having completed my Data Science program with WorldQuant University (WQU) and continued building my skills in Python, SQL, Data Analytics, and other areas of technology, I have been looking for opportunities to take what I have learned beyond theory and gain more practical experience.</p><p>This is one of the reasons I was particularly excited about DataraFlow.</p><p>DataraFlow is a structured and intensive program focused on Data Science, Machine Learning, and Generative AI, with an emphasis on learning, practical assignments, collaboration, and continuous problem-solving.</p><p>My goal is to become a well-rounded Data Scientist who can use data and technology to tackle real-world problems and develop solutions that create meaningful impact. So, when I came across this program, I saw it as an opportunity to strengthen the foundation I already have, learn new things, and gain more exposure to practical ways of applying my knowledge.</p><p>This is a program that I highly anticipated, and in just Week One (1), I can say I am not disappointed at all.</p><p>The first week has been a combination of setting things up, learning, practicing, solving problems, completing assignments and assessments, and getting familiar with how the program works. And although it is only the beginning, I already feel like I am on the right path.</p><h3>A Well Organized Beginning</h3><p>One thing that immediately stood out to me was the orderliness and coordination of the team. Everything seems to be well structured, and there is a clear direction for what we are expected to do.</p><p>We were also placed into different groups, with each group having a mentor. I have only introduced myself in my group so far, and I am looking forward to becoming more involved, contributing to discussions, learning from others, and hopefully contributing something meaningful to the group as we progress.</p><p>I believe that being surrounded by people who are also learning, asking questions, solving problems, and sharing ideas will be an important part of this journey.</p><h3>Setting Things Up</h3><p>We started the week by carrying out some basic installations and environment setups.</p><p>This involved preparing our computers for the work ahead by setting up tools such as Python and VS Code, creating a virtual environment, installing the necessary Python packages, and getting our development environment ready for the internship.</p><p>At first, these things can seem like simple technical requirements, but I realized that having a properly configured environment is actually an important part of programming and data science.</p><p>We worked with tools and packages that will be useful throughout the program, including libraries for data analysis, visualization, and machine learning, as well as tools for working with notebooks and managing our development environment.</p><p>There were a few moments where I had to troubleshoot some issues during the setup, but getting through them was also part of the learning process.</p><p>It reminded me that working in technology isn’t always about everything going perfectly the first time. Sometimes, you have to investigate an issue, try different solutions, and figure out what went wrong.</p><h3>Getting My Hands Dirty With Python</h3><p>We also had a lot of tasks, assignments, and assessments to complete.</p><p>This was probably one of my favorite parts of the week.</p><p>I was very proud of myself tackling some of the problems using Python. Some of the questions looked simple at first, but actually solving them required me to think about the logic behind the code.</p><p>During the week, I worked on different Python concepts and problems involving:</p><p><em>Variables and data types<br>User input<br>if, elif, and else statements<br>for and while loops<br>Lists, tuples, sets, and dictionaries<br>Functions<br>Local and global variables<br>== and is<br>File handling<br>Python exceptions<br>Modules<br>String manipulation<br>Palindromes<br>FizzBuzz<br>Fibonacci sequences<br>Factorials<br>Guessing games<br>Word counting<br>Finding the largest and second-largest numbers<br>Working with even numbers<br>Building simple calculators</em></p><p>What I particularly enjoyed was that I wasn’t just reading about these concepts. I was actually writing code and trying to solve problems with them.</p><p>For example, learning about loops became much clearer when I had to use them to generate Fibonacci numbers, create multiplication tables, count words, and build a guessing game.</p><p>Working with dictionaries also became more practical when I used them to represent information such as a student’s details or items in a shopping cart.</p><p>I also got to understand some concepts that I had previously found confusing. One example was the difference between == and is. I initially thought they were almost the same, but I now understand that == checks whether two values are equal, while is checks whether two variables refer to the same object.</p><p>These might look like small concepts, but understanding them properly makes a difference when writing and debugging code.</p><h3>More Than Just Writing Code</h3><p>One thing this first week reminded me of is that learning programming is not just about memorizing syntax.</p><p>It is about learning how to approach a problem.</p><p>Sometimes I knew the Python syntax but didn’t immediately know how to put everything together to solve the problem. Other times, I wrote something that looked right but didn’t produce the result I expected.</p><p>That process of making mistakes, understanding why something didn’t work, correcting it, and trying again is becoming an important part of my learning.</p><p>I am beginning to understand that the goal isn’t to write perfect code on the first attempt.</p><p>The goal is to understand the problem, break it down, and keep working until you find a solution.</p><p>And I believe this way of thinking will become even more important as we move into more advanced areas of Data Science and Machine Learning, where the problems will not always have straightforward answers.</p><h3>Looking Back at Week One</h3><p>Looking back, I would say Week One was a combination of setup, learning, practice, problem-solving, and getting familiar with the DataraFlow environment.</p><p>There is still a lot ahead, and I know the difficulty will increase as we progress through the program.</p><p>But if Week One is anything to go by, I believe there is a lot to look forward to.</p><p>I am especially excited about moving beyond the basics and applying what I am learning to more practical Data Science, Machine Learning, and Generative AI problems.</p><p>Having already built a foundation through my Data Science studies at WQU, I see DataraFlow as another opportunity to challenge myself, strengthen my skills, and gain more exposure to solving problems in practical settings.</p><p>I know there will be difficult moments ahead. There will probably be more bugs, more confusing concepts, more assignments that make me question my life 😂, and definitely more things to learn.</p><p>But that’s part of the journey.</p><p>For now, I am grateful for the opportunity to learn, improve my Python skills, interact with other learners, and be part of a structured learning environment.</p><p>I am looking forward to seeing how much I will learn, how much I will improve, and most importantly, how these skills will prepare me to tackle real-world problems as I continue my journey toward becoming a Data Scientist.</p><p>One week down, many more to go. 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2932a3197940" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>