<?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 Khurram on Medium]]></title>
        <description><![CDATA[Stories by Khurram on Medium]]></description>
        <link>https://medium.com/@khurram.rukhsar?source=rss-bcd91d49dfad------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*1EbIumZg36NJ_t30uSvyTA.jpeg</url>
            <title>Stories by Khurram on Medium</title>
            <link>https://medium.com/@khurram.rukhsar?source=rss-bcd91d49dfad------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 21:23:32 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@khurram.rukhsar/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[Code Red — You Are Probably Angry Because of Colours]]></title>
            <link>https://medium.com/@khurram.rukhsar/code-red-you-are-probably-angry-because-of-colours-61ace73c65de?source=rss-bcd91d49dfad------2</link>
            <guid isPermaLink="false">https://medium.com/p/61ace73c65de</guid>
            <category><![CDATA[india]]></category>
            <category><![CDATA[mental-health]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Khurram]]></dc:creator>
            <pubDate>Mon, 25 Aug 2025 17:04:58 GMT</pubDate>
            <atom:updated>2025-08-25T17:04:58.746Z</atom:updated>
            <content:encoded><![CDATA[<h3>Code Red — You Are Probably Angry Because of Colours</h3><p>“Sight” is one of the five senses of humans. The shape, the texture, the structure of an object, it all affects our perception of the object when viewed by our eyes. But there is one more attribute that every object holds that plays part in our perception of it — colour. Colours are fascinating, they induce a wide range of emotions in a human mind when light reflects after striking their surface.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Y4IOfwkCnuN4zJ4A" /><figcaption>Photo by <a href="https://unsplash.com/@mariogogh?utm_source=medium&amp;utm_medium=referral">Mario Gogh</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>It has been established by <a href="https://journals.sagepub.com/doi/full/10.1177/2158244014525423">research</a> that colours hold psychological significance and can directly influence our mood, attention, and even physiological responses. Certain colours when presented to us trigger certain emotional responses. They hold the power to alter our emotional state. For instance, warm colours like red, orange, and yellow are known to heighten alertness and evoke feelings of urgency, but prolonged exposure can also lead to agitation, anxiety, and mental fatigue.</p><p>When it comes to media, the production houses spend a great deal of time and money to design the most eye catching set with the most attention grabbing combination of colours paired with flashy texts and transition animations. All for one goal — to hook the viewer. In a rat race of TRP ratings, the goal of educating people with news and information is lost. Instead, the show with the most flashy animations, popping colours and a panel shouting at each other is considered to be the holy grail of digital media.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/739/1*HApI1ne-3hNy7lsC7tCUdQ.jpeg" /></figure><p>This was quite evident during the Pakistan-India war of May 2025. Indian media went viral all around the world during the tensions and for all the wrong reasons. People were aghast at seeing a panel of 12 people screaming at each other with anchor not even trying to moderate, there were instances where anchors literally were literally shouting at the dozen “experts” they had invited. <a href="https://www.dawn.com/news/1911728">There was war mongering, chest thumping, fake news, sensationalism and what not.</a></p><p>But beyond the noise, the visuals themselves were a weapon. The screens were soaked in red — scrolling tickers, urgent banners, and breaking news labels flashing every few seconds. Red is not just a colour on screen; it is a signal to our brains to be on guard, to prepare for danger. Psychologists have long shown that red stimulates the fight-or-flight response, elevates heart rate, and narrows focus. In other words: the colour alone was enough to make audiences feel anxious or angry, even before they processed the words being spoken.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*NBzCVegm2XDMu3U4zlTwJw.jpeg" /></figure><p>Curious about how deep this effect goes, I decided to put this to the test. I picked an Indian TV show most infamous for its fake news and sensationalism, <a href="https://www.youtube.com/watch?v=pPJ2xDmOA7E">Debate with Arnab — Republic TV</a>. It’s a daily show so I picked the recording for 7th May — the day when India launched unprovoked strikes on Pakistan — a 55 minutes long show.</p><h4>Methodology:</h4><p>Using computer vision techniques, I wrote a Python program to break down video footage frame by frame and measure the percentage of each colour on screen. For better intuition I decided to analyze only 9 colours i.e., VIBGYOR colour and black and white colours too. I used K-Means clustering to cluster the pixel colours on screen into one of the given 9 colours.</p><p>What I found was unsurprising yet alarming: red consistently dominated the visual palette during the most heated debates. Instead of balanced colour schemes, the broadcasts leaned into warm, aggressive tones designed to amplify tension.</p><h4>Results:</h4><p>The following gif shows the colour distribution of the screen per frame for the given talk show. It shows how Red dominates most of the time.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*ayeSpY6vlUlrmVscxDhAEg.gif" /><figcaption>Color Distribution Per Frame</figcaption></figure><p>This is not accidental. Media houses know that overstimulation — flashing graphics, clashing colours, and loud transitions — holds attention. But the cost is psychological. Viewers walk away not more informed, but more agitated. The “Code Red” of our times isn’t just about breaking news; it’s about how the very design of the screen manipulates our emotions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EbI79gQVEHSI0xWTwkQUBg.png" /></figure><p>The pie chart makes this even clearer: while green, violet, and other calmer colours barely register, <strong>red and black together consume the majority of visual space</strong>. In design theory, red is tied to urgency and alarm, black to seriousness and gravity. Together, they create a psychological “weight” that keeps viewers locked in fight-or-flight mode.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/558/1*OBxFtK4UGNLJ9WiebXC6uA.png" /></figure><h4><strong>Key Observations:</strong></h4><ul><li><strong>Red dominates</strong> the visual composition of the video, accounting for more than <strong>80%</strong> of the total frame time. This suggests a highly intense and emotionally charged visual theme.</li><li>Secondary dominance comes from <strong>Black (7.62%)</strong> and <strong>Orange (5.07%)</strong>, which often accompany red in conveying urgency, conflict, or drama.</li><li>Cooler tones like <strong>Indigo (3.54%)</strong> and <strong>Blue (0.67%)</strong> appear occasionally, possibly during calmer or transitional segments.</li><li>Neutral shades <strong>White (2.00%)</strong> and <strong>Black (7.62%)</strong> provide visual contrast but remain minor compared to red.</li><li><strong>Green, Yellow, and Violet</strong> together account for less than 1% of the frame time, indicating minimal representation of nature, optimism, or creative tones.</li></ul><h4>Discussion:</h4><p>The dominance of <strong>red</strong> highlights the deliberate use of a strong colour tone, likely aimed at evoking emotional intensity and capturing viewer attention. The relatively minor presence of cooler or neutral tones suggests the video maintains a high level of dramatic or urgent atmosphere throughout.</p><p>From a content analysis perspective, such a colour distribution can inform:</p><ul><li><strong>Media studies</strong>: understanding the emotional narrative of the video.</li><li><strong>Marketing and advertising</strong>: aligning brand visuals with desired psychological impact.</li><li><strong>Computer vision research</strong>: validating automated methods for scene classification based on colour profiles.</li></ul><p>So the next time you sit down to watch the news, pay attention not just to what is being said, but to what is being shown. Ask yourself: is the anger you feel truly about the issue, or is it a colour trick being played on your brain? Awareness is the first step to breaking the cycle. Because if we don’t, we risk living in a perpetual state of outrage — all thanks to a shade of red.</p><h4>Conclusion:</h4><p>This analysis demonstrates how simple frame-wise colour dominance metrics can reveal meaningful insights into a video’s emotional and aesthetic composition. The method is scalable, interpretable, and easily adaptable to other visual datasets, making it a valuable tool in both academic research and industry applications.</p><h4>Let’s Connect:</h4><p>Thanks for reading!<br>If you’re interested in the code or want to explore the visuals further, check out the full project on GitHub:</p><p>Github: <a href="https://github.com/KhurramRukhsar/NewsColourAnalysis">https://github.com/KhurramRukhsar/NewsColourAnalysis</a></p><p>And if you’d like to connect professionally, here’s my LinkedIn: <a href="https://www.linkedin.com/in/khurramrukhsar/">https://www.linkedin.com/in/khurramrukhsar/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=61ace73c65de" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quit Being So Negative]]></title>
            <link>https://medium.com/@khurram.rukhsar/quit-being-so-negative-c90f207673e2?source=rss-bcd91d49dfad------2</link>
            <guid isPermaLink="false">https://medium.com/p/c90f207673e2</guid>
            <category><![CDATA[mental-health]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[politics]]></category>
            <category><![CDATA[entrepreneurship]]></category>
            <dc:creator><![CDATA[Khurram]]></dc:creator>
            <pubDate>Sun, 20 Jul 2025 08:06:51 GMT</pubDate>
            <atom:updated>2025-07-20T08:08:41.392Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0iifSfyYAEW_Nx_9ZcImrQ.avif" /></figure><blockquote>Growing up, we didn’t have cable TV at home or a computer, we just had a good old antenna picking up state television (Pakistan Television — PTV). That meant that we only had two channels to browse through all day. As kids, only relevant media was the 15 minutes of cartoon segment by a morning show just before going to school, some 20 minutes of cartoons again at around 5 PM, a drama at 8 PM with everyone huddled around the TV and Khabarnama (9 PM News Bulletin).</blockquote><p>The world back then didn’t seem that gloomy. Maybe it was the innocence of childhood, or maybe it was just that we had restricted access to information. It really is true, ignorance is bliss. The other day I was having a discussion with a friend about the consumption of media and its effects. We both agreed that over the years, generally, people have gotten angrier. I have no data or source to back this claim but just my own observation. My hypothesis was that the news reports are generally negative around us and consumption of its 24/7 news coverage, augmented by thrilling social media posts and clickbait YouTube vlogs have pushed the people to their limits.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*nVeDox2PEPgek3XD" /><figcaption>Photo by <a href="https://unsplash.com/@libraryofmedicine?utm_source=medium&amp;utm_medium=referral">National Library of Medicine</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Now the question was “Is the news really more negative?” or was I just throwing a statement out there without any evidence to back it up with. I did what I always do, I went down a rabbit hole and found that I am not the only one who feels like it. <a href="https://www.bbc.com/future/article/20140728-why-is-all-the-news-bad">Multiple people</a> have voiced their <a href="https://www.theguardian.com/commentisfree/2018/feb/17/steven-pinker-media-negative-news">concerns</a>, in a much better and scientific way. But I wanted to confirm whether it was true for my case — whether Pakistani news around me was more negative or not. I wanted to gauge the sentiment of the news but not by me, not subjective, I wanted it to be objective. And nothing’s more objective than a machine. Thus, I started a small project and decided to use machine learning — I decided to do sentiment analysis of the news.</p><p>I had 3 goals for this project listed below:</p><ol><li>Confirm if there’s a bias in the sentiment distribution using the Chi-Square Test.</li><li>Visualize and see what kind of bias there is.</li><li>Are the title sentiments matching the content sentiment? If not then how is the mismatch distributed.</li></ol><p>This project had a data pipeline built that scraped the newspapers almost daily and then all data was cleaned, aggregated and analysed —<strong> <em>5144 news articles taken from two of the top English newspapers of Pakistan for over 30 days</em>.</strong></p><p><strong><em>NOTE: Following paragraph has technical details, you can skip it if you want to.</em></strong></p><p>I picked up two of the top English newspapers of Pakistan and I webscraped their daily newspapers using BeautifulSoup on Google Colab. I did it for several days, 27 to be precise. I compiled all the news articles in the newspapers into Pandas dataframes with titles and body of the articles separated into different columns. I combined tools from both the NLTK and Hugging Face libraries. I used NLTK’s sentence tokenizer to break the text into individual sentences, allowing for more granular sentiment evaluation. For the actual sentiment detection, I utilized the powerful transformers library from Hugging Face, specifically the pre-built pipeline for sentiment analysis, with the model set to ‘Twitter Roberta Base Sentiment’. This model was chosen because it classifies the sentiment into three categories, i.e. Positive, Negative and Neutral. Neutral was the most important category as all journalism is supposed to be neutral. This combination of tokenization and sentiment analysis enabled me to efficiently analyze the emotional tone of text data at the sentence level.</p><p>For the body, I got back sentiment of each chunk created after tokenization. I simply saw which sentiment was dominating in all the chunks and classified the news as such.</p><blockquote>What makes this issue even more concerning is the psychological weight headlines carry. Numerous studies in media psychology have shown that people tend to remember headlines more vividly than the body of an article — even when the content contradicts the headline. This cognitive shortcut, coupled with our tendency to skim rather than read deeply, means that a misleadingly negative title can shape public perception long before the facts are processed. I wanted to test this theory by removing my own bias from the equation and letting machine learning models objectively analyze the sentiment landscape.</blockquote><h4>Results:</h4><h4>Chi-Square Test:</h4><p>We performed Chi-Square Test for both the title sentiment distribution and content sentiment distribution. We got following results:</p><p>Title Sentiment:- Chi-square statistic: 3323.359, p-value: 0.000</p><p>Content Sentiment:- Chi-square statistic: 3660.714, p-value: 0.000</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/736/1*NTXKgIICYT0FBMNsYww9BA.jpeg" /></figure><p>The p &lt; 0.05 clearly indicates that there is an imbalance of sentiment in both the titles and content. Now we needed visualizations to see what kind of imbalance there was.</p><h4>Heatmaps:</h4><p>Here are some interesting results but before looking at them, one thing to keep in mind is that <strong>the issue at hand is not the negative news per se, but the uneven distribution of sentiment and the apparent mismatch between the sentiment of the title/headline of a news and that of the content of the news itself.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/548/1*OJDdZ2j0I6a1jO2E2gngnw.png" /></figure><p>The heatmap shows that for the most part title and content of news are in sync as you can see the central tile being the darkest, representing that out of 5144 news articles observed, 2723 were the ones which had a neutral sounding title and the content inside was neutral too. The diagonal is what ideal journalism should be i.e., the title and content in sync.</p><p>But we do notice that a large chunk is taken up by the tiles where title and content of a news are not in sync. <strong>One must note that the journalistic integrity demands the title and content’s sentiment to match</strong>, even if the title is positive when the content is neutral or negative, that too goes against journalistic standards. Now, we need to analyze the mismatched instances. The figure below shows only that.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/539/1*InByxmGc6vDX-edvzpL5rw.png" /></figure><h4>Bar Plots:</h4><p>We can see that amongst the group of mismatched title vs content, the top spot is taken up by the kind of news where the title is negative but the content was in fact rather neutral. This is one of the classic clickbait examples where the title is often sensationalized with a negative sentiment to force the reader to click on it whereas the news itself had no real controversy, urgency, or negativity upon closer inspection.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/540/1*PVmyE5UO4RKXviGNwRd67A.png" /></figure><p>When we look at which kind of titles has been mismatched the most, we see that the negative titles are the most mismatched. That means that in majority of the instances the title was negative, whereas the news content particularly wasn’t. It is notable that the share of positive titles’ sentiment which mismatched their content’s is considerably low showing a clear bias towards making the titles more negative and clickbait worthy.</p><p>The mismatched content bar graph tells the similar story and is self explanatory when seen with the context of mismatched title bar graph.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/549/1*rGR4p2dtnmFgZVTWgVrboQ.png" /></figure><h4>Venn Diagrams:</h4><p>When we dive a bit deeper and look at the data grouped by sentiment more intuitively in the form of Venn diagrams, we see how less of an overlap there is between the title and content for the positive and negative sentiment whereas for neutral sentiment, the overlap is greater than the non-overlapping area.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dn9tBGCM37jzE6-LmW69qA.png" /></figure><p>For positive sentiment see that the content was more positive whereas titles weren’t on par. For negative sentiment, we see a contrast, the title takes the bigger chunk for this sentiment and the content follows second. This contrasting behaviour is a pattern and a clear indication that negative titles are encouraged more as they act as clickbait or even as a ragebait for the news outlets.</p><h4>Overall Sentiment Match vs Mismatch:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/590/1*ee6cLl3WGUBSVsjFrqx8YA.png" /></figure><p>When we look at the overall percentages of match vs mismatch, we see a whopping 60.8% of news did not have the same sentiments when it came to their title and content.</p><h4>Sunburst Chart:</h4><p>To completely see the flow and distribution of sentiments from title to content, we have this really powerful sunburst chart. This plot bundles up all the disparities into one figure.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*32SWR2APEbuc00izuy81-A.png" /></figure><p>To conclude, I must reiterate, that it’s not the negative news which I am questioning, of course the world is not rainbows and butterflies where nothing wrong ever happens. The question is of the disparity between the headlines and the actual news, between the mismatch, and how that mismatch is skewed towards negative sentiment. This ragebaiting affects the reader’s opinion as in this day and age of scroll-culture where users prefer consuming micro-doses of content, a headline would make far more impressions than the actual news content. Print media is only a reflection of digital media, I am sure if such an analysis is done for the digital media, the results wouldn’t be much different.</p><h4>Let’s Connect:</h4><p>Thanks for reading!<br>If you’re interested in the code or want to explore the visuals further, check out the full project on GitHub:<br>GitHub: <a href="https://github.com/KhurramRukhsar/NewspaperSentimentAnalysis">https://github.com/KhurramRukhsar/NewspaperSentimentAnalysis</a></p><p>And if you’d like to connect professionally, here’s my LinkedIn: <a href="https://www.linkedin.com/in/khurramrukhsar/">https://www.linkedin.com/in/khurramrukhsar/</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c90f207673e2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Data Analysis Life Cycle]]></title>
            <link>https://medium.com/@khurram.rukhsar/data-analysis-life-cycle-d238d3d96564?source=rss-bcd91d49dfad------2</link>
            <guid isPermaLink="false">https://medium.com/p/d238d3d96564</guid>
            <category><![CDATA[tableau]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Khurram]]></dc:creator>
            <pubDate>Sun, 27 Nov 2022 11:49:13 GMT</pubDate>
            <atom:updated>2022-11-27T11:49:13.771Z</atom:updated>
            <content:encoded><![CDATA[<p>Data analysis is the process of collecting data and then analyzing that data to extract meaningful insights so that those insights can help the stakeholders in taking necessary decisions and solving problems.</p><p>Data analysis life cycle consists of six phases in total.</p><ul><li>Ask</li><li>Prepare</li><li>Process</li><li>Analyze</li><li>Share</li><li>Act</li></ul><h3><strong>Ask</strong></h3><p>You start the process of data analysis just like how you start the process of solving any problem in your daily life i.e. by asking yourself what is the ‘To Find’ section.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*0neHSv10bCJNErF5" /><figcaption>Photo by <a href="https://unsplash.com/@epicantus?utm_source=medium&amp;utm_medium=referral">Daria Nepriakhina 🇺🇦</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>By pinpointing what is asked of you, you can gather only the relevant data later. It’s like you have a sea of data at your behest but you need only a bucket of it for the job, so in the ‘Ask’ phase you define what data you need to be put into that bucket.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Rbw95702-s05GTCI" /><figcaption>Photo by <a href="https://unsplash.com/es/@brooksrice?utm_source=medium&amp;utm_medium=referral">Brooks Rice</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>You take the stakeholders on board and clearly define the questions that need to be answered. This is the foundation of data analysis; if this goes wrong, the whole building of analysis comes crumbling down later. Therefore, the analyst should be clear, complete, concrete, and concise in his communication with the stakeholders.</p><h3>Prepare</h3><p>In the ‘Preparing’ stage, we gather, store, and organize the data that is needed to solve the questions that were defined in the ‘Ask’ phase. The data could be taken from an established data warehouse, from a separate database or new data can be created as per the needs.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xtu7Xl5KDotH0lQZ" /><figcaption>Photo by <a href="https://unsplash.com/@mbaumi?utm_source=medium&amp;utm_medium=referral">Mika Baumeister</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Then that data is stored and decisions are taken about what form it should be stored in and what kind of privacy it needs to have. Should the data be stored in some external storage device, on an internal storage device, or cloud. Lastly, the data is organized in a form that is best to use for the posed questions.</p><h3>Process</h3><p>The data that was gathered, stored, and organized in the ‘Prepare’ stage is brought to the ‘Process’ stage so that it can be cleaned. Data is bias prone, a person who is recording or collecting data may not realize it but their personal biases or demographic limitations can cause the data to be skewed. These biases are managed in the process stage.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ol_FbeQcqVzt9nH8" /><figcaption>Photo by <a href="https://unsplash.com/@purzlbaum?utm_source=medium&amp;utm_medium=referral">Claudio Schwarz</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Null values are another issue that is fixed in the Process stage. Often some attributes are not recorded/answered while collecting data and that observation is given a null value. These null values are assigned some numeric value that makes it possible to include them in the calculations without impacting the result.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/602/1*E1YMlWSYsNvWsm5JLkH1Ag.jpeg" /><figcaption>Outliers</figcaption></figure><p>In addition to all this, ‘outliers’ are another discrepancy that is dealt with in the process phase. Outliers are anomalies that occur because of certain factors. It could be abnormal conditions, a series of coincidences, or some instrument malfunction that results in those anomalies. It is a good practice to take care of them before analyzing the data so that the analysis is accurate.</p><h3>Analyze</h3><p>In the ‘Analyze’ phase the cleaned data is put into formulas and equations so that calculations can be performed. Spreadsheets or SQL are used to perform calculations. You can use any language that you’re comfortable with for this phase.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Ot7Z44va6GZOxYqP" /><figcaption>Photo by <a href="https://unsplash.com/@chrisliverani?utm_source=medium&amp;utm_medium=referral">Chris Liverani</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>One can use Python, R, SQL, or any other language that has the packages to do calculations. This is a very crucial phase as it is the basis of all the visualizations that you’re going to do later on. Therefore if you use a wrong formula or use the wrong function then it can potentially make your whole analysis null and void.</p><h3>Share</h3><p>Good communication is the key to being a good data analyst. A data analyst has to be a good storyteller. It means that they should be able to effectively communicate their findings to the stakeholders who might not be from a technical background.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*olGXVKaiBLBRL0J5MQMtIA.png" /><figcaption>A good storyteller captivates their audience</figcaption></figure><p>Therefore a data analyst has to find a way to make those stakeholders see the trends in their findings and make them understand the gravity of those findings. For that visualization is used.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*VhXhycwKAnA_oWJc" /><figcaption>Photo by <a href="https://unsplash.com/@lukechesser?utm_source=medium&amp;utm_medium=referral">Luke Chesser</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In the ‘Share’ phase bar charts, pie charts, graphs, heat maps, flowcharts, and other visualization means are used to transform facts into figures. Dashboards and reports are made to summarize all visualizations in one place. This phase is important as the decision power rests with the stakeholders and if they are not convinced by a compelling presentation of facts and figures then all the work an analyst has done becomes futile.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*EAFB5vX2FkwNb9eq" /><figcaption>Photo by <a href="https://unsplash.com/@campaign_creators?utm_source=medium&amp;utm_medium=referral">Campaign Creators</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>For visualization purposes, different tools and platforms can be used. Microsoft Excel, Power BI, Tableau, Python, and R can be used for visualization.</p><h3>Act</h3><p>‘Act’ is the last and final phase of the data analysis life cycle. In this phase, the stakeholders decide whether they are convinced by your findings and presentation or not.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*l1KJU29Ge_IgkbGx" /><figcaption>Photo by <a href="https://unsplash.com/es/@kaleidico?utm_source=medium&amp;utm_medium=referral">Kaleidico</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>They may ask you to redo some analysis, they might point out any mistakes or if the analysis lacks anything, they might not be convinced at all and reject all of your propositions. This phase decides whether you need to convince the stakeholders more or are they convinced already.</p><h3>About Khurram</h3><ul><li>LinkedIn Profile — l<a href="https://www.linkedin.com/in/khurramrukhsar/">inkedin.com/in/khurramrukhsar/</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d238d3d96564" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>