<?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 Jack “Reuben” Bibi on Medium]]></title>
        <description><![CDATA[Stories by Jack “Reuben” Bibi on Medium]]></description>
        <link>https://medium.com/@jbibi1296?source=rss-8f1649a21b3b------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*CwEL8-Ng2V3vhNpu.jpg</url>
            <title>Stories by Jack “Reuben” Bibi on Medium</title>
            <link>https://medium.com/@jbibi1296?source=rss-8f1649a21b3b------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 04:05:08 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@jbibi1296/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[What train is my track on?]]></title>
            <link>https://medium.com/@jbibi1296/what-train-is-my-track-on-1e280a38ed38?source=rss-8f1649a21b3b------2</link>
            <guid isPermaLink="false">https://medium.com/p/1e280a38ed38</guid>
            <category><![CDATA[efficiency]]></category>
            <category><![CDATA[projects]]></category>
            <category><![CDATA[twitter]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[transportation]]></category>
            <dc:creator><![CDATA[Jack “Reuben” Bibi]]></dc:creator>
            <pubDate>Tue, 30 Jul 2019 17:41:12 GMT</pubDate>
            <atom:updated>2019-08-01T14:00:12.593Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uRAIn9jvu7ILaXKc.jpg" /></figure><p>If you have ever walked into Penn Station, you know that it’s a nightmare. Thousands of people waiting around, eyes glued to the TV screens, waiting to see their train’s track number. As soon as a track number is put up on the board, there are two different reactions. Either you are thinking, “That’s my track! Nobody else matters, and I don’t care who I run over as long as I get to my track” or you are on the other side thinking “Why can’t that be my track? Now I am getting trampled, and I’m not gonna be able to see what track my train is going to be at”.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PfZrE_9rAV4a2l0N" /><figcaption>An average day at NY Penn Station</figcaption></figure><p>A few weeks ago, while waiting for my train’s track to be posted, using my undeniable charm, I got friendly with a former-conductor. After talking for around 10 minutes, he saw my restlessness to get a good seat on my train. He calmly asked, “What train are you taking?” to which I responded, “The 7:05 train towards Bay Head”. I immediately saw a grin start to form on his face as he checked his phone and said: “Go to track 1 NOW!”. I didn’t know what to do, was he punking me or was this legit? Rather than run and never see this man again, I decided to focus on how he was able to get this info so quickly. After a lot of praise and a little bit of begging, he ended up spilling the most sought after secret in Penn Station. (This was all on the condition that I don’t pass on his secret, and I will keep my word.) Over the next few weeks, I have been one of the first people on the train, with the best seat, all without the sweat marks!</p><p>After seeing many of my friends in aggravation, I decided to think of an easy way for them to see their track number, all without revealing any secrets. I reached out to them and wanted to hear their requests and suggestions. After a few days, I decided that the best way to reach out to them would be through Twitter. I chose Twitter because users can see past train information, and it gives users an option to receive notifications for all tweets.</p><p>The next step was to create a python script to pull the data and push it to Twitter in an easily-understood, user-friendly tweet. After about an hour, I had a perfect working pipeline that would grab new data and format it correctly for Twitter. (I cannot share/post the script due to my promise with the former-conductor.) I needed a service that would be able to run my python script in the cloud continuously, so I went with PythonAnywhere. I like PythonAnywhere because I have used them for some previous projects of mine and they were able to run my scripts for me at no additional cost.</p><p>Lastly, I used IFTTT to post the tweets for me. I implemented a line of code in my script to send a POST-request to IFTTT with the tweet. When this request was submitted, IFTTT would take the data from the request and send out a tweet with this information.</p><p>The twitter account is @Deal_Train. The name is because of the city where my friends and I live in New Jersey. This Program/Account was created solely for people who are traveling via the NJ Transit from NY Penn Station towards Long Branch / Bay Head.</p><p>PS: I will make my account private once it hits 100 followers to keep an element of secrecy.</p><p>Reach out to me if you would like me to create an account for a different line on the NJ Transit.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1e280a38ed38" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How much am I worth?]]></title>
            <link>https://medium.com/@jbibi1296/how-much-am-i-worth-d6ef85f76586?source=rss-8f1649a21b3b------2</link>
            <guid isPermaLink="false">https://medium.com/p/d6ef85f76586</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[web-scraping]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[towards-data-science]]></category>
            <category><![CDATA[salary-negotiations]]></category>
            <dc:creator><![CDATA[Jack “Reuben” Bibi]]></dc:creator>
            <pubDate>Fri, 10 May 2019 18:31:56 GMT</pubDate>
            <atom:updated>2019-05-13T14:25:47.891Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*T1WgYy-EnrbOEklpWkWqTQ.jpeg" /></figure><p>The hardest part of job hunting is when the recruiter turns to you and says “So, what kind of salary are you looking for?”</p><p>At this moment my heart dropped and I started to flip through the story of my life. “I won that award that time, plus $800”, “I fell off that slide when I was 3 years old, subtract $2k”, “I am 22 without a job, subtract $5k”, “I have not read more than 12 books in my life, subtract $10k”… I realized that when put on the spot, I can only think of the negative parts of my life instead of all the great things that I have done.</p><p>I realized that this would be a great project to create. A model that can take in either a resume or a job posting as an input and the output would be a salary or salary range.</p><p>I decided to use Google Jobs as my source of jobs. This is because they aggregate their jobs from a bunch of other job platforms. Additionally, they post salary estimates from ‘Glassdoor”, “Built in NYC”, and “PayScale”. <br>I downloaded a <a href="http://chromedriver.chromium.org/">chrome driver</a> to be used with <a href="https://www.seleniumhq.org/">Selenium</a>.</p><p>You can use this <a href="https://towardsdatascience.com/web-scraping-using-selenium-python-8a60f4cf40ab">blog post</a> by <a href="https://towardsdatascience.com/@bandiatindra">Atindra Bandi</a> to learn more about Selenium and how it can be used.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*35TlgaNibqIbfH99w67xyQ.png" /><figcaption>Inspecting how Google formats it’s job listings</figcaption></figure><p>After inspecting the Google job platform, I was able to identify which data I would pull from the job posting. I wanted the title, company name, job posting, location, and the estimated salary (From Glassdoor or the others).</p><p>I started by creating a function to take an input of a search term and Selenium would open up chrome and open a Google job search using the search term given.</p><pre><strong>def</strong> open_link(job_name):<br>    job_name = job_name.replace(&#39; &#39;,&#39;_&#39;)<br>    job_url_name = job_name.replace(&#39;_&#39;,&#39;+&#39;)<br>    url = f&#39;https://www.google.com/search?q=<strong>{job_url_name}</strong>&amp;ibp=htl;jobs#fpstate=tldetail&amp;htidocid=&#39;<br>    driver = webdriver.Chrome(executable_path=&#39;./chromedriver&#39;)<br>    driver.get(url)</pre><pre>open_link(str(input())</pre><p>Once I used Selenium to open up the browser and perform the search, I inspected and was able to identify how I would be instructing BeautifulSoup to pull the data that we want.</p><pre>def get_job_data():<br>    # Get HTML from the page<br>    soup = BeautifulSoup(driver.page_source, &#39;lxml&#39;)<br>        jobs = soup.find_all(&#39;li&#39;)<br>        job_list = []<br>    # Loop through each job and save the data to a list of dictionaries<strong><br>        for</strong> job <strong>in</strong> jobs:<br>            job_dic = {}<br>            job_dic[&#39;title&#39;]= job.find(&#39;h2&#39;,{&#39;jsname&#39;:&#39;SBkjJd&#39;}).text<br>            job_dic[&#39;company&#39;]=job.find(&#39;div&#39;,{&#39;class&#39;:&#39;pbHUre tcoBdd&#39;}).text<br>            job_dic[&#39;body&#39;]=job.find(&#39;span&#39;,{&#39;style&#39;:&#39;line-height:1.5em&#39;}).text<br>            job_dic[&#39;location&#39;]=job.find_all(&#39;div&#39;,{&#39;class&#39;:&#39;tcoBdd&#39;})[1].text<br>            <strong>try</strong>:<br>                job_dic[&#39;salary&#39;]=job.find(&#39;span&#39;,{&#39;class&#39;:&#39;zE8vH&#39;}).text.replace(&#39;CA&#39;,&#39;&#39;)<br>            <strong>except</strong>:<br>                job_dic[&#39;salary&#39;]=np.nan<br>            # Append the job dictionary to a list <br>            job_list.append(job_dic)<br>    # Create a Data Frame from the list<br>    df = pd.DataFrame(job_list)<br>    return df</pre><pre>job_dataframe = get_job_data()</pre><p>Now that we have the proper job data, we can continue to pull more jobs from different search words.</p><p>I created a function that can pull up to 150 jobs at a time but it will require a little bit of manual scrolling and input.<br>After running the function, you will be asked to enter a search word (no quotes required). Once the page is opened, scroll on the list of jobs until the page stops refreshing. Once you are done, go back to the script, and type a ‘y’ to proceed. The script will go through each job and pull the necessary data and it will return a DataFrame.</p><pre><strong>def</strong> manually_get_job_data():<br>    job_name = str(input(&quot;Enter Job Title&quot;)).replace(&#39; &#39;,&#39;_&#39;)<br>    job_url_name = job_name.replace(&#39;_&#39;,&#39;+&#39;)<br>    pre_url = f&#39;https://www.google.com/search?q=<strong>{job_url_name}</strong>&amp;ibp=htl;jobs#fpstate=tldetail&amp;htidocid=&#39;<br>    driver = webdriver.Chrome(executable_path=&#39;./chromedriver&#39;)<br>    driver.get(pre_url)<br>    proceed = 0<br>    <strong>while</strong> proceed != &#39;Y&#39;:<br>        proceed = str(input(&#39;Are You Ready To Proceed? (Y/N)&#39;)).upper()<br>    <strong>for</strong> i <strong>in</strong> [pre_url+link.attrs[&#39;id&#39;][4:] <strong>for</strong> link <strong>in</strong> BeautifulSoup(driver.page_source, &#39;lxml&#39;).find_all(&#39;div&#39;,{&#39;jsname&#39;:&#39;x5pWN&#39;})]:<br>        driver.get(i)<br>    soup = BeautifulSoup(driver.page_source, &#39;lxml&#39;)<br>    jobs = soup.find_all(&#39;li&#39;)<br>    job_list = []<br>    <strong>for</strong> job <strong>in</strong> jobs:<br>        <strong>if</strong> (job.find(&#39;h2&#39;,{&#39;jsname&#39;:&#39;SBkjJd&#39;}) != <strong>None</strong>) <strong>and</strong> (len(job.find_all(&#39;div&#39;,{&#39;class&#39;:&#39;tcoBdd&#39;}))&gt;1):<br>            job_dic = {}<br>            job_dic[&#39;title&#39;]= job.find(&#39;h2&#39;,{&#39;jsname&#39;:&#39;SBkjJd&#39;}).text<br>            job_dic[&#39;company&#39;]=job.find(&#39;div&#39;,{&#39;class&#39;:&#39;pbHUre tcoBdd&#39;}).text<br>            job_dic[&#39;body&#39;]=job.find(&#39;span&#39;,{&#39;style&#39;:&#39;line-height:1.5em&#39;}).text<br>            job_dic[&#39;location&#39;]=job.find_all(&#39;div&#39;,{&#39;class&#39;:&#39;tcoBdd&#39;})[1].text<br>            <strong>try</strong>:<br>                job_dic[&#39;salary&#39;]=job.find(&#39;span&#39;,{&#39;class&#39;:&#39;zE8vH&#39;}).text.replace(&#39;CA&#39;,&#39;&#39;)<br>            <strong>except</strong>:<br>                job_dic[&#39;salary&#39;]=np.nan<br>            job_dic[&#39;avg_salary&#39;]=avg(job_dic[&#39;salary&#39;])<br>            job_list.append(job_dic)<br>    new = pd.DataFrame(job_list)</pre><pre>    <strong>return new<br>df = manually_get_job_data()</strong></pre><p>After pulling over 6,000 jobs, it was time to go through the data and clean it up.</p><p>Our data had 2 issues:<br>1. Only about 3,000 columns had estimated salary data<br>2. The estimated salary data was a range of values (EX: $75k — $120k)</p><p>In order to fix this, I dropped all rows that did not have a salary, and I created a function to go through each cell in the salary column and find the average salary for the range (EX: $75k — $120k = $97,500). Once it got all of the averages, I created a new column in the DataFrame to store all of these new numbers.</p><p>In order to run text through a statistic model, we would have to turn our text into numbers.<br>The cleansing I used for the text consisted of tokenizing, lemmatizing, removing stop words, and vectorizing.<br>I decided to use SkLearn’s TFIDF-Vectorizer to turn the text into numbers based on the frequency of the word in the document, and the frequency of the word in our entire corpus (all of our documents).</p><p>They hyper-parameters that I decided to use for the vectorizer was</p><pre>n_grams = (1,3),<br>max_df = .85<br>min_df = .15<br>binary = True</pre><p>Now that the data was ready to be put through a stats model, I performed a train_test_split on it and initially ran it through a LinearRegression. This was to get a “base score” that I can use to see the accuracy of my other models, and also, for inference.<br>I went on to use Ridge, Lasso, Random Forest, Gradient Boost, and even a Neural Network.</p><p>For all of these models, my metric was Root Mean Square Error.<br>Glassdoor’s normal range ± $32,214<br>LinearRegression — $30,595<br>Lasso — $30,458<br>Ridge — $30,560<br>Random Forest — $19,290<br>Gradient Boost — $18,379<br>Neural Network — $28,736</p><p>The model predicts salaries with a lower error that Glassdoor!</p><p>I then decided to place my <a href="https://www.jackbibi.dev/p/my-resume.html">resume</a> inside of the model to see what it predicted.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gO4Xr5PRCBiMl8Fj2IJGLg.png" /><figcaption>My model’s prediction my estimated yearly salary (with margin of error above it)</figcaption></figure><p>This is a pretty accurate estimation for entry level jobs as a Data Scientist / Data Analyst in New York City.</p><p>Thank you for reading!</p><p>P.S. I am continuously working on this project to make it more accurate. I recently performed a lot of feature engineering and vectorized the location and the job title. The current error that I am receiving is around $14,500! <br>I will post soon with my updates and a link to my github to learn more about my project.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d6ef85f76586" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[All or Nothing]]></title>
            <link>https://medium.com/@jbibi1296/all-or-nothing-754e8e14e385?source=rss-8f1649a21b3b------2</link>
            <guid isPermaLink="false">https://medium.com/p/754e8e14e385</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[transitioning]]></category>
            <category><![CDATA[short-story]]></category>
            <dc:creator><![CDATA[Jack “Reuben” Bibi]]></dc:creator>
            <pubDate>Tue, 07 May 2019 17:43:43 GMT</pubDate>
            <atom:updated>2019-05-07T17:43:43.333Z</atom:updated>
            <content:encoded><![CDATA[<p>Growing up, my parents always taught me to never miss out on an opportunity. Well, boy were they smart! <br>In January of 2019, I received an email from a cousin of mine, Jack Hidary, sent an email blast to over 100 relatives inviting us to come take a tour of Google’s offices in NYC a month. At this point in my life, I had been working in a family business and we were almost ready to launch a new division which I had been working on for over a year. I knew that I needed to spend as much time as possible making sure the launch is successful and I didn’t have time to go on tours. On the other hand, how often can an ordinary person get a semi-personal tour of one of the greatest companies in the world! After losing a night of sleep on this thought, I sent my RSVP and went to work that morning with a clear head.</p><p>Fast forward to February 4th 2019, the day has finally come. I’ll never forget the progression of my attitude that day. At first, I was so excited to go to and I was smiling ear to ear. As the day went on, the grin started to fade along with the excitement, as soon as it was time to leave the office, I was convinced that I would be more productive by staying in the office and getting my work done. I heard my mother in my head telling me to “seize the opportunity” and it gave me the motivation to put my computer to sleep and to go to Google. What she didn’t know was that I had my laptop in my briefcase so I can get as much work done as possible throughout the tour.</p><p>As I walked out of the elevator, I felt like I was home yet I’ve never been here before. The combination of forward-thinking, strong work ethic, and priority on innovation, was prevelant all around and I need to be a part of it. After we were given a tour, Jack spoke to all of us explaining the future of technology and how important it is for us to get involved. I have known this for years, but any attempt at studying a field in the tech sector has been thwarted by my parents saying “will you be able to support a family?”. Now was my chance to jump in, and I needed Jack to be the one to convince my parents to support me through it.</p><p>I spoke to Jack for a bit about AI and Machine Learning and he sensed an interest in me and refused to let it go to waste. Jack said “you should think about taking a bootcamp to learn more about Data.” I looked to my right, made eye contact with my parents, and slyly replied “I would love to but it’s up to my parents.” As soon as my dad heard this, he replied with his template answer “will you be able to support family?”. I pointed at Jack who then replied “Data Science is one of the highest paying jobs right now.” As my father was trying to think of way to rebuttle, my mother stepped in and asked her favorite cousin “how do we sign him up?” I could not believe what I had just heard. “Was this really happening?” “NOW?”</p><p>What followed was one of the most important car rides of my life. My parents were in the front and I was in the back. We were researching the bootcamp available, arguing about how I would leave my job, planning on how we would make sure that my responsibilities in the business would be taken care of, and ended with me thanking them both for supporting me in taking this huge step in my life.</p><p>If there is one thing to take out of this story, it is to never pass on an opportunity.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=754e8e14e385" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>