<?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[Commute Chronicles: Decoding Transportation - Medium]]></title>
        <description><![CDATA[A series of short research articles that dive deep into everything transit — public and private. Transit systems are crucial urban lifelines that impact societies, commerce, and the environment. We analyze the data to get some perspective and explore solutions to inefficiencies. - Medium]]></description>
        <link>https://medium.com/commute-chronicles-decoding-transportation?source=rss----3e067d8d9094---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Commute Chronicles: Decoding Transportation - Medium</title>
            <link>https://medium.com/commute-chronicles-decoding-transportation?source=rss----3e067d8d9094---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 16:21:51 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/commute-chronicles-decoding-transportation" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[California High-Speed Rail: A High-Level Perspective]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/california-high-speed-rail-a-high-level-perspective-875b8270046c?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/875b8270046c</guid>
            <category><![CDATA[rails]]></category>
            <category><![CDATA[public-transit]]></category>
            <category><![CDATA[transportation]]></category>
            <category><![CDATA[california]]></category>
            <category><![CDATA[high-speed-rail]]></category>
            <dc:creator><![CDATA[Advait]]></dc:creator>
            <pubDate>Wed, 02 Aug 2023 22:41:22 GMT</pubDate>
            <atom:updated>2023-08-02T22:45:59.183Z</atom:updated>
            <content:encoded><![CDATA[<p>The planned California High-Speed Rail Project encompasses more than 800 miles of rail and 24 stations. A central objective of this initiative is to establish a rail connection between the major urban hubs of Los Angeles and San Francisco, with a travel time of under 3 hours. The project is currently in the construction phase and the commencement of initial train operations is not expected before 2030. This piece will concentrate on offering a geographical overview of the project, while also drawing comparisons to the existing Amtrak Tracks in California. The analysis is supported by data furnished by the California High-Speed Rail Authority (CHSRA). As usual, you can find a link to my Python code at the conclusion of this article. I intend for this article to be part of a series discussing this massive public transit project.</p><p>The California High-Speed Rail initiative is divided into two distinct stages. The initial phase, Phase 1, encompasses the construction of a rail route that will link San Francisco to Los Angeles, traversing several cities within the Central Valley. Moving on to Phase 2, there are two planned extensions of the main line: one to the north, extending from Merced to Sacramento, and another to the south, extending the line from Los Angeles to San Diego.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*mJthxJSuR7XNtNz6.png" /><figcaption>Image 1: Planned California High Speed Rail Routes and Stations, for both the phases.</figcaption></figure><p>The initial phase of the proposed California High-Speed Rail is partitioned into ten distinct sections, as depicted on the accompanying map. Presently, the ongoing construction activities are predominantly concentrated within the Central Valley segments of the route, spanning from Merced to Bakersfield.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*FQITj350VvliENVL.png" /><figcaption>Image 2: Sections of Phase 1 of California’s High Speed Rail project</figcaption></figure><p>The second phase comprises a pair of segments, the first entailing an expansion from Merced to Sacramento, while the second involves a southern extension stretching from Los Angeles to San Diego.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*QaVI3QznlScBhS2f.png" /><figcaption>Image 3: Sections of Phase 2 of California’s High Speed Rail project. Section 1 is in Red.</figcaption></figure><p>I will now analyze the intended path of the California High-Speed Rail in contrast to the current Amtrak route operational in California. It’s important to highlight that the proposed high-speed rail aims to connect San Francisco and Los Angeles in a mere 3 hours, whereas the quickest Amtrak rail journey from Emeryville (the nearest station to San Francisco) to Los Angeles takes over 12 hours. Also note that the current Amtrak route covers a much larger area within California compared to the planned High Speed Rail Network.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/989/0*KH4fHngWdyVHTafs.png" /><figcaption>Image 4: Side-by-side Comparisons between planned CA High Speed route and current CA Amtrak routes</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*MvAKTFbxx2DXQYwB.png" /><figcaption>Image 5: The existing Amtrak Routes in California (highlighted in orange) superimposed with the intended route of the upcoming California High-Speed Rail (highlighted in blue).</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*03A3JrqVG2SPNqWB2QlyzQ.png" /><figcaption>Image 6: Amtrak’s Coastal Starlight, the exclusive direct train route connecting the Bay Area and Los Angeles, requires more than 12 hours of travel time. “EMY” stands for Emeryville and “LAX” stands for Los Angeles. Source: <a href="https://www.amtrak.com/tickets/departure.html">https://www.amtrak.com/tickets/departure.html</a></figcaption></figure><p>I trust that this article has provided you with valuable insights and expanded your understanding of the California High-Speed Rail plan, as well as its comparison with the current Amtrak Network. If this has captured your interest, you may also find my article on the ownership of Amtrak’s rail tracks in California intriguing.</p><p>Link to article on the ownership of Amtrak’s rail tracks in California: <a href="https://medium.com/commute-chronicles-decoding-transportation/who-owns-amtraks-rail-lines-in-california-cb28d61fa3ad">https://medium.com/commute-chronicles-decoding-transportation/who-owns-amtraks-rail-lines-in-california-cb28d61fa3ad</a></p><p>Sources:</p><ol><li>My Kaggle Notebook containing Python code used to create visualizations: <a href="https://www.kaggle.com/code/adyg1234/california-high-speed-rail-analysis">https://www.kaggle.com/code/adyg1234/california-high-speed-rail-analysis</a></li><li>California High Speed Rail planned routes: <a href="https://gis.data.ca.gov/datasets/83492c31c5604917856580447ab09f76/explore">https://gis.data.ca.gov/datasets/83492c31c5604917856580447ab09f76/explore</a></li><li>California High Speed Rail planned stations: <a href="https://gis.data.ca.gov/datasets/california-high-speed-rail-statewide-stations/explore">https://gis.data.ca.gov/datasets/california-high-speed-rail-statewide-stations/explore</a></li><li>“<em>California high-speed rail will connect the mega-regions of the state, contribute to economic development and a cleaner environment, create jobs and preserve agricultural and protected lands. The system will run from San Francisco to the Los Angeles basin in under three hours at speeds capable of over 200 miles per hour.</em>”: <a href="https://hsr.ca.gov/high-speed-rail-in-california/">https://hsr.ca.gov/high-speed-rail-in-california/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=875b8270046c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/california-high-speed-rail-a-high-level-perspective-875b8270046c">California High-Speed Rail: A High-Level Perspective</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[BART ridership: Unpacking some temporal trend]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/bart-ridership-unpacking-some-temporal-trend-efceb63e2261?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/efceb63e2261</guid>
            <category><![CDATA[time-series-forecasting]]></category>
            <category><![CDATA[transportation]]></category>
            <category><![CDATA[transit]]></category>
            <category><![CDATA[bay-area]]></category>
            <dc:creator><![CDATA[Abhishek Roy]]></dc:creator>
            <pubDate>Sat, 29 Jul 2023 21:07:51 GMT</pubDate>
            <atom:updated>2023-07-29T04:39:46.204Z</atom:updated>
            <content:encoded><![CDATA[<h3>BART ridership: Unpacking some temporal trends</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/0*0sA7RKiv-r4A-1KH.jpg" /><figcaption>Source: <a href="https://www.bart.gov/guide">https://www.bart.gov/guide</a></figcaption></figure><p>Like every other transit system across the US, COVID-19 impacted the Bay Area Rapid Transit (BART) system heavily. However, the impact seems to be far more severe in the case of the BART as compared to many other states across the US. With work-from-home policies taking a stronghold in the work force, many people just don’t need to commute to work anymore. In this article, let us dive deeper into the BART ridership trends that we saw in the previous article. We will examine the general shape and trend of ridership pre-COVID, use some basic time-series modelling to predict what kind of ridership levels would we have seen had it not been for COVID-19, and finally get an idea of how long it might take for current ridership levels to recover back to pre-COVID levels.</p><h3>I. Getting a sense of the general ridership trend pre and post COVID-19</h3><p>Let us first revisit the chart from the previous article that shows the cumulative monthly entries from 2001–2022.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F13.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F13%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/532650acac4be9fd47a6fa72bc359757/href">https://medium.com/media/532650acac4be9fd47a6fa72bc359757/href</a></iframe><p>As we can see in Figure 1, in the part of the chart before the sudden drop after COVID, there seem to be two underlying patterns — one seasonal pattern that repeats every 12 months and another pattern that seems to have an upward trend, peaking every 8 odd years. The shorter 12 month seasonal pattern can be attributed to lower BART traffic during January-December (holiday season) and peak BART traffic during September-October (peak work and tourist season). But the larger 8–9 year oscillating trend may be due to a number of socio-economic reasons. Let us look at one possible reason — changes in population. Check out these two charts:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F15.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F15%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/6ef32f33fafcbced59330f732fdbe0a3/href">https://medium.com/media/6ef32f33fafcbced59330f732fdbe0a3/href</a></iframe><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F32.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F32%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/4198305049b5bd618a14c3087c67de8a/href">https://medium.com/media/4198305049b5bd618a14c3087c67de8a/href</a></iframe><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F19.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F19%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/7e5a31551012f6d5dc63722f46052ab4/href">https://medium.com/media/7e5a31551012f6d5dc63722f46052ab4/href</a></iframe><p>Figure 2 and 3 do not readily seem very comparable, but broadly, we can see that while population is increasing in Figure 2, ridership is also increasing in Figure 3. Similarly, when the population trend starts plateauing in 2016, ridership also seems to fall after 2016. Let us compare growth rates for both values and overlay them.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F21.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F21%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/cb80357b5e4e1927f8cd1594cdc957a2/href">https://medium.com/media/cb80357b5e4e1927f8cd1594cdc957a2/href</a></iframe><p>Here it seems clearer that as population growth rates start to fall, BART ridership growth rates also fall. So a potential explanation behind BART’s longer term trends could be fluctuations in population.</p><h3>II. Using basic time series modelling to see what ridership trends would have been if not for COVID-19</h3><p>In order to do this time series analysis, I have essentially followed these:</p><ul><li>Examine time series from 2001–2018 Feb<br>- Take First Difference<br>- Check trend and seasonality<br>- Examine ACF and PACF plots</li><li>Train model on data from 20001 - 2018 February</li><li>Test on data from 2018 March — 2020 March</li><li>Retrain model on data from 2001–2020 February and Project from 2020 March — 2022 March</li></ul><p>The specific details, charts, and the parameters of the SARIMA model (Seasonal Autoregressive Integrated Moving Average) used can be found in the Kaggle link attached at the end of the article. For now, let us look at how the model performs on the test data visually:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F23.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F23%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/9fd3ab68efc2e65b26f4e5e719af1ee5/href">https://medium.com/media/9fd3ab68efc2e65b26f4e5e719af1ee5/href</a></iframe><p>The model seems to be working decently well. The red line looks like it is capturing most of the variation in the time series. Let us now look at how the model predicts ridership patterns might have evolved in an alternate world where COVID did not happen.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F25.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F25%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/a93a1d1011ded717eb1aace4bc4edea3/href">https://medium.com/media/a93a1d1011ded717eb1aace4bc4edea3/href</a></iframe><p>As we see in Figure 7, ridership patterns might have been on a slightly downward trend, probably because of the general population patterns we saw earlier. Even then, ridership seems to be around the 400,000 levels. After COVID, the real ridership values are nearer to the 130,000 levels, which is a huge difference in people using BART every month.</p><h3>III. Checking post-COVID trend and estimated time to reach pre-COVID levels</h3><p>Finally, let’s do a very basic analysis to get a sense of how much time it might take for current BART ridership levels to get back to pre-COVID levels. We’ll observe that the trend is more or less linear. Assuming work from home policies, perhaps ridership trends will not increase in an exponential fashion any time soon. We will use this linear trend to project into the future and get a rough idea of the long drawn out impacts of COVID on this transit system.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F34.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F34%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/cbe4e905a1aa7c416e1429df113b702f/href">https://medium.com/media/cbe4e905a1aa7c416e1429df113b702f/href</a></iframe><p>The shorter 12 month seasonal trends still seem to be present. Once we remove the seasonal component from Figure 8 as well, we will be left with a pretty linear trend. I will overlay a line with a approximated slope of this linear trend to show just how linearly BART ridership seems to be increasing through the end of 2022.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F36.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F36%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/f8b755a675d9a3d87b6f7d1917a470b0/href">https://medium.com/media/f8b755a675d9a3d87b6f7d1917a470b0/href</a></iframe><p>As we can the, the underlying trend for ridership recovery seems to be a linear line with a slope of about 4155, meaning an additional 4155 people are using BART every month. Using this information, we can extrapolate the ridership trend into the future and get a visual idea of how long it might take for ridership levels to reach 400,0000 (pre-COVID levels).</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F38.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F38%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/99bcfcbd761ccd7f0f10c11cff5850e7/href">https://medium.com/media/99bcfcbd761ccd7f0f10c11cff5850e7/href</a></iframe><p>As suggested by Figure 10, a linearly increasing trend might imply that it could possibly take until the beginning of 2028 to recover back to pre-COVID levels of ridership. Of course, this is a very simplistic model and it is very difficult to show when and how exactly ridership levels might recover, since there are many variables to consider. This model also uses data only until the end of 2022, so 2023 ridership might show slightly different recovery patterns. But hopefully, this gives some perspective into the deep effect of COVID on BART ridership levels.</p><p>Sources:</p><p>Kaggle notebook: <a href="https://www.kaggle.com/code/abhishekroy2000/commuter-chronicles-bart-time-series">https://www.kaggle.com/code/abhishekroy2000/commuter-chronicles-bart-time-series</a></p><p>Population data: <a href="https://data.census.gov/">https://data.census.gov/</a></p><p>Ridership data: <a href="https://www.bart.gov/about/reports/ridership">https://www.bart.gov/about/reports/ridership</a></p><p>BART image source: <a href="https://www.bart.gov/guide">https://www.bart.gov/guide</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=efceb63e2261" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/bart-ridership-unpacking-some-temporal-trend-efceb63e2261">BART ridership: Unpacking some temporal trend</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Who owns Amtrak’s Rail Lines in California?]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/who-owns-amtraks-rail-lines-in-california-cb28d61fa3ad?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/cb28d61fa3ad</guid>
            <category><![CDATA[california]]></category>
            <category><![CDATA[public-transit]]></category>
            <category><![CDATA[rails]]></category>
            <category><![CDATA[amtrak]]></category>
            <category><![CDATA[transportation]]></category>
            <dc:creator><![CDATA[Advait]]></dc:creator>
            <pubDate>Fri, 28 Jul 2023 19:39:47 GMT</pubDate>
            <atom:updated>2023-07-28T19:43:16.076Z</atom:updated>
            <content:encoded><![CDATA[<p>Amtrak, the national passenger company of the United States, offers services across 46 states and boasts an extensive network of rail routes, connecting over 150 destinations within California. What’s intriguing is that while Amtrak operates its own trains, it does not have ownership of the tracks it utilizes in the state. This prompts us to delve into the inquiry of who is responsible for owning and maintaining the rail infrastructure that Amtrak relies on in California, and what implications arise from this arrangement.</p><p>In this article, I will present visualizations to gain a better understanding of Amtrak’s operations in California and shed light on the actual owners of the tracks on which it runs. As always, you can find the link to my Python code for creating these visualizations at the end of this article.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/600/0*RVzOLGtFOjucaFON" /><figcaption>Image 1: Amtrak’s Coastal Starlight Line, (<a href="https://www.amtrak.com/san-francisco-bay-area-northern-california-train-routes">https://www.amtrak.com/san-francisco-bay-area-northern-california-train-routes</a>)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*AnoNXibfwQ3MaCVn.png" /><figcaption>Image 2: The rail tracks utilized by Amtrak in California, depicted in a darker shade of blue.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*_CYizqdVCe4_z0Jn.png" /><figcaption>Image 3: Depicting some of Amtrak’s significant routes in California, these lines span across a considerable portion of the State.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*8wg2R-xrl6fJ2L_h.png" /><figcaption>Image 4: Amtraks’s California Stations represented by blue dots. Note the higher concentration of dots near San Francisco Bay Area and the Los Angeles Area</figcaption></figure><p>The data utilized in my analysis is sourced from the “California Rail Network” dataset, which can be accessed through the California State Geoportal. This dataset comprises essential variables pertaining to each rail track segment, and I employ this information for conducting subsequent analyses detailed below.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*q1UDt2kwim892A7-.png" /><figcaption>Image 4: “Right of Way” ownership of Amtrak’s California Routes</figcaption></figure><p>The map presented above illustrates the “Right of Way” ownership of Amtrak’s California Routes. In this context, the entities highlighted in the map’s legend are the owners of the tracks, granting them priority over Amtrak’s trains. Amtrak, in turn, engages in negotiations with these parties to utilize these routes, often requiring payment of a fee. However, due to limited viable rail track alternatives, these negotiations may not always be advantageous for Amtrak. Notably, as indicated in the legend, Amtrak does not own any of the tracks on which its trains operate. Now, let’s delve deeper into the parties that have ownership of these tracks!</p><p>The five parties that own Amtraks’ rail tracks in California are:</p><ol><li>BNSF (Burlington Northern Santa Fe). BNSF Railway is one of the major freight railroad companies in the United States, operating an extensive network of tracks across various states</li><li>SCRRA (Southern California Regional Rail Authority). SCRRA is a regional commuter rail system serving the Southern California region with “Metrolink” trains.</li><li>SDNR (San Diego Northern Railway<strong>). </strong>SDNR is a regional commuter rail system serving the Southern California region with “COASTER” trains.</li><li>SMV (Santa Maria Valley Railroad). It is a short 14.8 mile railroad owned by Coast Belle Rail Corporation.</li><li>UP (Union Pacific Railroad). UP is one of the major freight railroad companies in the United States. It operates an extensive rail network covering several states and plays a crucial role in transporting goods and commodities across the country.</li></ol><p>Let’s now explore the ownership of Amtrak’s rail tracks in California in greater detail. The map clearly shows that a significant majority of Amtrak’s California routes are controlled by two for-profit companies: BNSF (Burlington Northern Santa Fe) and UP (Union Pacific Railroad). Both of these companies primarily focus on operating freight services on their tracks.</p><p>Union Pacific, also known by its ticker symbol “UNP,” is a publicly traded company listed on the New York Stock Exchange. On the other hand, Burlington Northern Santa Fe is under the ownership of Berkshire Hathaway, the conglomerate company led by Warren Buffet. This alignment of ownership has notable implications for Amtrak’s operations within the state. BNSF and UP, being primarily freight railroad companies, primarily prioritize and focus on operating freight transportation services on their tracks. As a passenger rail service, Amtrak shares these tracks with the freight operators, resulting in certain challenges related to scheduling, maintenance, and potentially slower or less flexible passenger train services, as freight trains often receive priority on the shared tracks.</p><p>Among the other major owners of Amtrak’s California tracks are SCRRA and SDNR, both of which are state-owned organizations with a primary focus on passenger service. This difference in ownership, compared to private freight companies like BNSF and UP, may result in fewer challenges for Amtrak as they share the tracks, as passenger service providers often align their priorities more closely with Amtrak’s operations.</p><p>Let us now look at who all operate freight service on the Amtrak routes in California that are owned by Union Pacific.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*Q1wu8RYLWzy1qkee.png" /><figcaption>Image 5: Freight service Operators on Amtrak routes in California that are owned by Union Pacific</figcaption></figure><p>As depicted in the map, Union Pacific (UP) operates as the exclusive freight operator on most of its railroads where Amtrak operates. Nevertheless, in the areas highlighted in Purple and Pink colors, particularly around the Bay Area and Sacramento, UP shares freight operations with other companies like BNSF and CFNR (California Northern Railroad). Notably, this region includes the popular “Capital Corridor” commute service operated by Amtrak.</p><p>I conduct a similar analysis to investigate the freight service operators on the Amtrak routes in California that are owned by BNSF.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/688/0*IvvJ2pofmlcTH4vE.png" /><figcaption>Image 6: Freight service Operators on Amtrak routes in California that are owned by BNSF</figcaption></figure><p>Similar to the previous map, BNSF functions as the sole freight operator on the majority of its railroads in California where Amtrak operates. However, in the Central Valley Region, represented by the Grey and Light Blue colors, BNSF shares freight operations with other companies such as UP (Union Pacific) and CCT (Central California Traction Company).</p><p>I trust you found this article informative, shedding light on the ownership of the tracks that Amtrak operates on in California and its implications. Stay tuned for more engaging articles on transit-related topics. Thank you for reading!</p><p>Sources:</p><ol><li>My Kaggle Notebook containing Python Code used for Analysis : <a href="https://www.kaggle.com/code/adyg1234/california-amtrak-rail-network-analysis">https://www.kaggle.com/code/adyg1234/california-amtrak-rail-network-analysis</a></li><li>California Rail Network Data: <a href="https://gis.data.ca.gov/datasets/2ac93358aca84aa7b547b29a42d5ff52/explore">https://gis.data.ca.gov/datasets/2ac93358aca84aa7b547b29a42d5ff52/explore</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cb28d61fa3ad" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/who-owns-amtraks-rail-lines-in-california-cb28d61fa3ad">Who owns Amtrak’s Rail Lines in California?</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[San Francisco Uber Mobility Data Analysis]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/san-francisco-uber-mobility-data-analysis-8ae6d43a9f0a?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/8ae6d43a9f0a</guid>
            <category><![CDATA[maps]]></category>
            <category><![CDATA[public-transit]]></category>
            <category><![CDATA[san-francisco]]></category>
            <category><![CDATA[geospatial]]></category>
            <category><![CDATA[clustering]]></category>
            <dc:creator><![CDATA[Advait]]></dc:creator>
            <pubDate>Mon, 24 Jul 2023 03:05:45 GMT</pubDate>
            <atom:updated>2023-07-26T20:54:38.398Z</atom:updated>
            <content:encoded><![CDATA[<p>Over the past decade, ridesharing giants like Uber and Lyft have experienced remarkable growth, solidifying their position as dominant players in the intra-city transit market. This surge in ridesharing usage has drawn the interest of transportation planners, who are eager to examine these trends in order to pinpoint potential locations for developing public transit infrastructure as an alternative solution. Although obtaining data on their service usage is often challenging due to their private nature, Uber has recently launched “Uber Movement.” This platform provides insights into Uber ridership trends in specific major cities, offering valuable information on travel times, speeds, and mobility patterns.</p><p>This article aims to analyze mobility data through Python code and apply basic machine learning techniques to group commute areas into transit zones. For more detailed information and access to the code notebook on Kaggle, please refer to the provided link at the end of this article.</p><p>Please be aware that the data utilized in this analysis corresponds to Quarter 1 of 2020, and the traversals represent the average total number of rides between 10:00 am and 6:00 pm during that specific time frame.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZELhCbHsNBd4iwX7WiVEXg.png" /><figcaption>Image 1: First 5 rows of Uber Mobility Dataset for San Francisco</figcaption></figure><p>As depicted in Image 1, the mobility dataset contains only four variables. The geographic area where the traversals take place is represented by the “wktGeometry” variable.</p><p>Below, you can find a plot of one instance of the geographical area represented by the “wktGeometry” variable. It’s important to note that all instances of this variable represent hexagonal geographical areas.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/562/0*zIGg8FXiLCutjjhX.png" /><figcaption>Image 2: Plot of an instance of the “wktGeometry” variable.</figcaption></figure><p>Below is a basic plot showing all the geographical spots represented in the San Francisco Uber Mobility Pattern datase. I have included a labelled map of the San Francisco area for your reference.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/437/0*dJsMN6awIwz5s6NL.png" /><figcaption>Image 3: Plot of all Geographical Area represented in SF Mobility Data</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/556/0*_DuMazD78BUIXbrw" /><figcaption>Image 4: San Francisco Area Map from <a href="https://www.nps.gov/safr/planyourvisit/maps.htm">https://www.nps.gov/safr/planyourvisit/maps.htm</a></figcaption></figure><p>After performing basic exploratory data analysis, it becomes evident that the majority of geographical areas in the dataset exhibit an average of fewer than 200 Uber-ride traversals.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/591/0*LVki9kHMi6RDe7xR.png" /><figcaption>Image 5: Histogram Plot of the “traversals” variable</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/832/1*Mqph_p-az5_F_DjotOQ_LA.png" /><figcaption>Image 6: Number of observations belonging to each bucket in “traversals” variable</figcaption></figure><p>In order to concentrate the analysis on areas with a substantial volume of Uber rides, I set a threshold of 200 traversals as an indication of high rideshare traffic within those regions. Subsequently, I plotted the areas with more than 200 traversals on a map of San Francisco, using marker sizes proportional to the number of traversals in each area. Larger dots on the map represent higher Uber ride traversals, signifying busier zones with greater rideshare activity.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/802/0*BCxsHoBPpKW5gkku.png" /><figcaption>Image 7: Plot of Areas in San Francisco with &gt; 200 Uber “traversals” on a daily basis, marker size is proportional to the number of traversals in that area</figcaption></figure><p>After plotting the areas in San Francisco with high Uber traversals, the next step is to strategically group these regions to make it easier to create public transit stops as a practical alternative. One way to achieve this is by employing the weighted K-Means clustering algorithm, which is an unsupervised machine learning method. To put it simply, K-Means is used to group objects in space into “K” clusters, with the goal of minimizing the distance between each point and its closest cluster center. I may consider writing an article later to provide a more detailed explanation of the K-Means algorithm, but in summary, it helps in organizing data points efficiently by forming clusters based on their spatial relationships.</p><p>Below is a map showing the results of using the K-Means algorithm to cluster the high Uber traversal Zones into 100 clusters. For your reference, there are around 117 Muni Metro stops and close to 2,500 Muni Bus stops in San Francisco.</p><p>Presented below is the map showcasing the results of geospatial clustering accomplished through the weighted K-Means algorithm. In this process, the high-Uber traversal areas have been grouped into 100 distinct clusters. It’s important to note that the “weight” attributed to each point during clustering corresponds to the number of Uber traversals passing through the hexagon, which the point is a centroid of.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/802/0*sCIYBokPo0_a1AWq.png" /><figcaption>Image 8: The high Uber traversal areas in San Francisco are organized into 100 “clusters.” Each cluster is represented by points of the same color, and these points are surrounded by points of different colors.</figcaption></figure><p>I now repeat the same analysis as above, but now grouping the high Uber traversal areas into 1,000 cluster instead!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/802/0*6nCTmd_CsB0Hg_M5.png" /><figcaption>Image 9: The high Uber traversal areas in San Francisco are organized into 1,000 “clusters.” Each cluster is represented by points of the same color, and these points are surrounded by points of different colors.</figcaption></figure><p>To gain perspective, we can compare the aforementioned plots with the map displaying SFMTA bus and light rail stops, aiming to identify possible gaps in public transit accessibility. I previously presented these maps in a prior Medium article, which you can find linked at the end of this current article. However, it’s essential to bear in mind that the transit options considered in this comparison do not encompass services like the Golden Gate Transit system and the Golden Gate Park shuttle system.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/790/0*1GrZkJRj4N8qTcIr.png" /><figcaption>Image 10: Updated map featuring the bus and light rail stops of the San Francisco Municipal Transportation Agency (SFMTA).</figcaption></figure><p>I trust you will find this article enlightening, and I am delighted to share my perspective on how the optimization of public transit stops can be achieved using the weighted K-Means clustering algorithms.</p><p>Source:</p><ol><li>My Kaggle Notebook: <a href="https://www.kaggle.com/code/adyg1234/sf-uber-mobility-data-analysis">https://www.kaggle.com/code/adyg1234/sf-uber-mobility-data-analysis</a></li><li>My earlier Medium Article on using Geopandas to plot SFMTA transit routes and stops: <a href="https://medium.com/commute-chronicles-decoding-transportation/plotting-san-francisco-muni-smta-routes-using-geopandas-67409f4646fc">https://medium.com/commute-chronicles-decoding-transportation/plotting-san-francisco-muni-smta-routes-using-geopandas-67409f4646fc</a></li><li>Uber Mobility Heatmap data for San Francisco: <a href="https://movement.uber.com/explore/san_francisco/mobility-heatmap/">https://movement.uber.com/explore/san_francisco/mobility-heatmap/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8ae6d43a9f0a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/san-francisco-uber-mobility-data-analysis-8ae6d43a9f0a">San Francisco Uber Mobility Data Analysis</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[BART ridership: Some high level overview and perspective]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/bart-ridership-some-high-level-overview-and-perspective-6d78f6785db0?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/6d78f6785db0</guid>
            <category><![CDATA[transit]]></category>
            <category><![CDATA[transport]]></category>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[bay-area]]></category>
            <dc:creator><![CDATA[Abhishek Roy]]></dc:creator>
            <pubDate>Fri, 21 Jul 2023 06:30:26 GMT</pubDate>
            <atom:updated>2023-07-21T06:28:12.602Z</atom:updated>
            <content:encoded><![CDATA[<p>The Bay Area Rapid Transit, or “BART”, is an arterial network of trains running through San Francisco and connecting major locations across the East Bay. Its lines stretch as far as Richmond, Antioch, Dublin, and North San Jose. Well in advance of me coming to the Bay Area for university, I was made aware of the importance of BART for getting from point A to point B across the Bay. 4 years later, it’s the reason I reach my office in Oakland from my home in San Francisco in 25 minutes. Most people around the area are well aware of the indispensability of BART, but lack awareness of the general ridership trends of the network. How many people does BART service every month? Do people use BART more and less in certain months? To what extent has COVID impacted ridership? What are the most popular stations? Let’s see if ridership data can answer these questions.</p><h3>BART Network Map</h3><p>Here is a map of the BART stations from BART’s website:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*sKuob_X7B_Ts3o9C.png" /></figure><p>As you can see, it connects all major hubs from San Francisco to East Bay. You may notice a gap between Millbrae (where the yellow line ends) and San Jose. Most locations there are serviced by the Caltrain, which is another rail line.</p><h3>Overall Ridership Trends</h3><p>To analyze ridership, I first downloaded the entry-exits matrices from BART’s website link, attached in the Sources at the end. The data was cleaned and appended to form two datasets — total entries per month per Station and total exits per month per Station. These datasets were used for the following charts and tables. Let’s first look at the general Entry/Exit trends.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F5.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F5%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/5094ac25b5cfa37b80dd92dbb76e90d2/href">https://medium.com/media/5094ac25b5cfa37b80dd92dbb76e90d2/href</a></iframe><p>The aggregate trend lines for Entries and Exits across all stations on weekdays is almost exactly the same for both Entries and Exits. What we can infer from this chart is that before the pandemic, from 2015–2020, there were more than 400,000 people using the BART monthly on weekdays. That number has dipped down sharply during the pandemic and has slowly recovered around 150,000. This shows the long lasting effects of the pandemic and enduring work-from-home policies that have reduced public transport ridership.</p><p>There is also some visible periodicity and seasonality in the graph. Just eyeballing the crests and troughs show that ridership generally dips during December-January and peaks during September-October. Here is a zoomed view at 2015–2019.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F9.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F9%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/e709d768fc7d5e7cb0d3b8f6db03aa51/href">https://medium.com/media/e709d768fc7d5e7cb0d3b8f6db03aa51/href</a></iframe><p>To prove that the periodicity of the data is annual, I made a quick autocorrelation graph. Autocorrelation, simply put, is a way to measure the similarity of a variable’s value with delayed values of itself or “lags”. In the graph below, the lag with the highest autocorrelation is 12, meaning the periodicity of the data is 12 months.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F11.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F11%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/8d5cfc1d245fd17b7a1c2a6f720fb17e/href">https://medium.com/media/8d5cfc1d245fd17b7a1c2a6f720fb17e/href</a></iframe><h3>Station Popularity</h3><p>Let us now get a sense of the popularity of each station. For this, we simply sum up all the weekday entries by station for two years prior and post the pandemic, i.e., 2018–2019 and 2021–2022. For 2018–2019, these are the top 10 stations:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ESOe2TAjuXAnwFkpNZyOPw.png" /><figcaption>Total Ridership from 2018–2019 for each BART station</figcaption></figure><p>And for 2021–2022, these are the top 10 stations:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KAg4troxZyof-oVngIvzUA.png" /><figcaption>Total Ridership from 2021–2022 for each BART station</figcaption></figure><p>This gives us a sense that some stations might have had a slower recovery compared to other stations. 12th Street Oakland, for example, dropped from #4 to #8 in the rankings. 19th Street Oakland fell below the top 10, while the top 4 stations in San Francisco maintained their rankings — Embarcadero, Montgomery Street, Powell Street, Civic Center. The next chart helps visualize these trends and compare the volume across each station.</p><p>Yearly ridership trends by Top 10 stations</p><p>In the chart attached below, we can see the total yearly weekday entries for the top 10 stations according to total ridership in 2022. Notice that overall, the rankings of the stations don’t seem to have changed much after COVID, but the four busiest stations in San Francisco seem to have a steeper recovery curve after COVID than the rest of the stations.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplotly.com%2F%7EAbhishekRoy3141%2F7.embed%3Fautosize%3Dtrue&amp;display_name=Plotly&amp;url=https%3A%2F%2Fchart-studio.plotly.com%2F%7EAbhishekRoy3141%2F7%2F&amp;image=https%3A%2F%2Fchart-studio.plotly.com%2Fstatic%2Fwebapp%2Fimages%2Fplotly-logo.8d56a320dbb8.png&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=plotly" width="600" height="400" frameborder="0" scrolling="no"><a href="https://medium.com/media/bf9cf93bfa5bb7afa928aa4640045a58/href">https://medium.com/media/bf9cf93bfa5bb7afa928aa4640045a58/href</a></iframe><p>In the next few articles, we’ll diver deeper into this ridership data and some other data sources as well, conducting some mapping, time series analysis, and supervised/unsupervised learning to see if we can find possible areas of improvement and solutions to inefficiencies. For now, I hope these charts and tables have given you some idea about the invaluable transit system that has been servicing the Bay Area for the last five decades — BART.</p><p><em>Sources:</em></p><p><strong>BART Ridership Reports</strong>: <a href="https://www.bart.gov/about/reports/ridership">https://www.bart.gov/about/reports/ridership</a></p><p><strong>BART Ridership Watch:</strong> <a href="https://www.bart.gov/news/articles/2020/news20200225">https://www.bart.gov/news/articles/2020/news20200225</a></p><p><strong>BART Station names:</strong> <a href="https://www.bart.gov/sites/default/files/docs/station-names.xls">https://www.bart.gov › docs › station-names</a></p><p><strong>BART Map:</strong> <a href="https://www.bart.gov/system-map">https://www.bart.gov/system-map</a></p><p><strong>Kaggle Notebook:</strong> <a href="https://www.kaggle.com/code/abhishekroy2000/commuter-chronicles-bart-trends-over-time">https://www.kaggle.com/code/abhishekroy2000/commuter-chronicles-bart-trends-over-time</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6d78f6785db0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/bart-ridership-some-high-level-overview-and-perspective-6d78f6785db0">BART ridership: Some high level overview and perspective</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Evolution of Travelling to Work in the USA]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/evolution-of-travelling-to-work-in-the-usa-91a36f37b7a6?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/91a36f37b7a6</guid>
            <category><![CDATA[public-transit]]></category>
            <category><![CDATA[usa]]></category>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[transportation]]></category>
            <dc:creator><![CDATA[Advait]]></dc:creator>
            <pubDate>Tue, 18 Jul 2023 05:00:53 GMT</pubDate>
            <atom:updated>2023-07-24T03:08:01.615Z</atom:updated>
            <content:encoded><![CDATA[<p>Following the COVID-19 pandemic, mass transit systems throughout the USA have experienced significant declines in ridership, leaving some facing financial challenges. The primary factors contributing to the decrease in US public transport ridership after COVID-19 include the surge in “work from home” arrangements and an increase in car usage. To gain deeper insights into the trends among various transportation modes Americans utilize for their commutes, I analyze data from the Bureau of Transportation Statistics.</p><figure><img alt="Image 1: Primary transportation mode of American workers from 2001–2021 (in Millions)" src="https://cdn-images-1.medium.com/max/767/1*G7MNi1fp8EhSgvYRBtZSMQ.png" /><figcaption>Image 1: Primary transportation mode to work for American workers from 2001–2021 (in Millions)</figcaption></figure><p>As depicted in Image 1, a vast majority of Americans, totaling more than 100 million individuals, rely on automobiles for their daily commutes to work. On the other hand, alternative modes of transportation, including Public Transportation, Bicycles, and Motorcycles, constitute only a small fraction of the total transportation usage over the period 2001–2021. While the aforementioned chart offers valuable insights into transportation patterns, presenting the data in a percentage format, as shown below, might offer a more intuitive perspective.</p><figure><img alt="Image 2: Primary transportation mode of American workers from 2001–2021 (in Percentages)" src="https://cdn-images-1.medium.com/max/767/1*DGBq6C78Hkplg-IYPCubbA.png" /><figcaption>Image 2: Primary transportation mode to work for American workers from 2001–2021 (in Percentages)</figcaption></figure><p>As illustrated in Image 2, the usage of automobiles, including the significant “Drives self” category, is showing a gradual decline as a means of commuting to work. To understand the factors driving this trend, it becomes crucial to examine the chart closely, as it indicates a rising percentage of workers who are opting to work from home. For this reason, it becomes valuable to take a closer look at transportation trends concerning modes that fewer than 20% of American workers rely upon.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/772/0*u6qtlPKQIOgZ5fC6.png" /><figcaption>Image 3: Select Transportation modes to work for American workers from 2001–2021 (in Percentages)</figcaption></figure><p>As displayed in Image 3, there is a noticeable rise in the percentage of the American workforce working from home, resulting in a decreased reliance on traditional transportation modes for commuting. Although the impact of the COVID-19 pandemic exacerbated this trend in 2020 and 2021, the shift towards remote jobs had been steadily increasing even before that time period. Additionally, it’s worth noting that public transit constituted approximately 5% of the chosen transit mode for American workers from 2001 to 2020, but this percentage declined by half to around 2.5% in 2021.</p><p>Several factors contribute to this decline in public transit usage, including the need for social distancing due to COVID-19 and a growing flexibility towards remote work. However, one pertinent question arises: How does the decline in public transit usage compare to the decline in driving?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/776/0*pPHiCMIFcfj3XfzT.png" /><figcaption>Image 4: Changes in usage of Select Transportation modes to work for American workers from 2001–2021</figcaption></figure><p>As depicted in Image 4, the decline in the proportion of Americans using public transit to commute to work is much more pronounced compared to the decline in driving. In 2021, the share of public transit usage experienced an almost 50% drop, which is considerably larger than the decrease in car usage for commuting. One plausible hypothesis for this trend is that the lack of social distancing measures in public transit during the pandemic played a significant role in discouraging workers from utilizing it.</p><p>I hope you found my article insightful! Please leave your thoughts</p><p>Sources:</p><ol><li>I used data from the Bureau of Transportation Statistics for my analysis: <a href="https://www.bts.gov/content/principal-means-transportation-work">https://www.bts.gov/content/principal-means-transportation-work</a></li><li>The Python Code for my analysis is located in this Kaggle link: <a href="https://www.kaggle.com/code/adyg1234/principal-means-of-transportation-in-the-usa/input">https://www.kaggle.com/code/adyg1234/principal-means-of-transportation-in-the-usa/input</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=91a36f37b7a6" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/evolution-of-travelling-to-work-in-the-usa-91a36f37b7a6">Evolution of Travelling to Work in the USA</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Plotting San Francisco Muni (SFMTA) Routes using GeoPandas]]></title>
            <link>https://medium.com/commute-chronicles-decoding-transportation/plotting-san-francisco-muni-smta-routes-using-geopandas-67409f4646fc?source=rss----3e067d8d9094---4</link>
            <guid isPermaLink="false">https://medium.com/p/67409f4646fc</guid>
            <category><![CDATA[public-transit]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[geography]]></category>
            <category><![CDATA[san-francisco]]></category>
            <category><![CDATA[bay-area]]></category>
            <dc:creator><![CDATA[Advait]]></dc:creator>
            <pubDate>Tue, 18 Jul 2023 05:00:32 GMT</pubDate>
            <atom:updated>2023-07-31T20:37:25.449Z</atom:updated>
            <content:encoded><![CDATA[<p>In this article, I show you how to use the Geopandas and other Python libraries to plot San Francisco Public Transit routes on a map. The San Francisco Municipal Transportation Agency (SFMTA, commonly referred to as “Muni”) is a department of the City and County of San Francisco responsible for the management of all ground transportation in the city. I use publicly available data, as part of the San Francisco Open Data project to plot the maps below. The link to my code in Kaggle is at the end of this article.</p><p>Some preliminaries below for importing necessary libraries</p><pre>import geopandas as gpd<br>import numpy as np<br>import pandas as pd<br>import shapefile as shp<br>import matplotlib.pyplot as plt<br>import seaborn as sns<br>import shapely.wkt<br>import contextily as ctx</pre><p>Next, I read a dataset containing the Routes of SF Muni using the Pandas Library.</p><pre>df_route = pd.read_csv(&quot;/kaggle/input/sf-open-data-transportation/Muni_Simple_Routes.csv&quot;)<br>df_route.head()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xO4455jliBJUKBY5B5O7JA.png" /></figure><p>What is the “shape” variable though? It is in fact a set of (latitude, longitude) points that indicates the actual route of the MUNI vehicle. The line joining the points indicated in an instance of the variable is how the route is defined. Below, I examine an instance of the “shape” variable for this dataset.</p><pre>df_route[&quot;shape&quot;][0]</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qfBVbSb_WN-Havrs4qs3kA.png" /></figure><p>Below, I convert the Pandas Dataframe into GeoPandas dataframe. Note that it is important to define the “geometry” parameter of the geopandas instance, which in this case is the “shape” variable of the raw data.</p><pre>gp_df_route = gpd.GeoDataFrame(df_route, geometry = df_route[&quot;shape&quot;].apply(shapely.wkt.loads))</pre><p>Below is a simple plot of the route data using the GeoPandas dataframe I created above.</p><pre>gp_df_route.plot()<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/521/0*zUsh2QFETgQVh08Z.png" /></figure><p>The initial plot above gives us an indication of the transit routes with respect to latitudes and longitudes. However, it is very unclear about how the routes correspond to the actual geography of the San Francisco Bay Area. To do so, I plot the routes above, with an underlying map using the “contextily” library.</p><pre>gp_df_route.crs = &quot;EPSG:4326&quot;<br>gp_df_route = gp_df_route.to_crs(epsg=3857)<br>ax = gp_df_route.plot(figsize=(10, 10), alpha=1.0, edgecolor=&#39;blue&#39;)<br>ctx.add_basemap(ax, source=ctx.providers.CartoDB.Positron)<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/782/0*gJ0YadyA0X_zO0og.png" /></figure><p>The map above looks much better when giving geographical context to the SF Muni Transit routes. I will now use the “SERVICE_CA” and plot the different types of services offered by the SFMTA, using a different color for each type. This involves only a minor modification to the code above, where we now plot by column = “SERVICE_CA” and include a legend to the plot as well.</p><pre>ax = gp_df_route.plot(column = &quot;SERVICE_CA&quot;, legend = True, figsize=(10, 10), alpha=1.0,  legend_kwds={&#39;loc&#39;: &#39;upper left&#39;})<br>ctx.add_basemap(ax, source=ctx.providers.CartoDB.Positron)<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/782/0*csjWwZ9D_rnrarYj.png" /></figure><p>Hope you find this article useful and best wishes for analyzing transit routes across the world! Please leave any thoughts, questions and ideas!</p><p>Sources:</p><ol><li>Author’s Kaggle Analysis Notebook: <a href="https://www.kaggle.com/code/adyg1234/sf-muni-route-stops-visualization-basic-version">https://www.kaggle.com/code/adyg1234/sf-muni-route-stops-visualization-basic-version</a></li><li>SF Open Data Transit Routes Data: <a href="https://data.sfgov.org/Transportation/Muni-Simple-Routes/9exe-acju">https://data.sfgov.org/Transportation/Muni-Simple-Routes/9exe-acju</a></li><li>SFMTA Website: <a href="https://www.sfmta.com/">https://www.sfmta.com/</a></li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=67409f4646fc" width="1" height="1" alt=""><hr><p><a href="https://medium.com/commute-chronicles-decoding-transportation/plotting-san-francisco-muni-smta-routes-using-geopandas-67409f4646fc">Plotting San Francisco Muni (SFMTA) Routes using GeoPandas</a> was originally published in <a href="https://medium.com/commute-chronicles-decoding-transportation">Commute Chronicles: Decoding Transportation</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>