When and where is the traffic of Chicago?

Time, location, and land use in Chicago

This dataset contains each trip that occurred from 2018 and its characteristics like start and end times, fares, start and end places (in a privacy-aware way), distance, among others. As expected this dataset is enormous to be analyzed using spreadsheet software, only the raw data of one month has 8.8 million records.

As noted before, the Transportation Network Providers (TNP) is the largest dataset, which contains every rideshare trip made from 2018. This analysis is focused on January 2020, which contains 8 804 875 trips (around 1.45GB). To aggregate the points given by TNP, we used the Community Areas (CA). This contains all 77 community areas in the city, so we can use CA to identify which one has the highest traffic, at which times, and how congested is. Finally, the Boundaries Zoning Districts (ZON) dataset contains the 12 287 delimitations of land zoning by type.

The hypothesis of the analysis is that traffic is highly correlated with time, which means that peaks times are around 8 am and 5 pm when most of the people are going to work/school, or leaving from there, respectively. Also, is expected the areas with traffic and its surroundings have a lower average speed.

To reduce the complexities of the model of analysis, we will focus only on the pickup places (or start points) of the trip as the unit of traffic. Also, we will disregard weekends (including Fridays) and non-geolocalized trips to have a more consistent conclusion. A final assumption is that we identify a jammed area as any area with a reduction of 20% of its average speed.

As expected during the early morning the probability of getting a trip that will be stuck in a traffic jam at that timeframe is relatively low (less than 0.4), however between 2 and 4 am, there is a high probability in the south — near Pullman and South Deering. 6 am is when all areas start to get red, at 7 am almost all places have a probability of traffic around 50%. Afterward, places like Garfield Park, Midway Airport, and O’Hare Airport concentrated the jams until 9 am. Except for specific areas, the probability remains below 40% until 3 pm when areas close to the O’Hare Airport concentrate the jams.

Regarding the speed variation, it mostly follows the pattern of the probability of traffic. From the speed point of view, between 7 am and 8 am, almost all areas suffer a reduced average speed of 10% or worse. For example, the average speed is 30% lower in the areas where the airports are located. One hypothesis is that areas with high traffic probability have lower average speed, but this is not always true. At 11 am, the area near CSX Intermodal 59th Street Yard has a traffic probability of 50%, but a speed greater than 30% than its average.

Using a spatial join technique, we can match the points of trips to the land zoning. Land uses like manufacturing and downtown have the worst reduction of speed compared to the average. As noted in the previous section, Pullman and South Deering located in the south are affected by traffic jams during the early morning. These areas have manufacturing land use. Likewise, central parts of Chicago that had more traffic probability all day, are located in downtown land use, which correlated with the table. In summary, land use seems to have implications with traffic and average speed.

Related posts:

Blossoming love during a springtime stroll Lost in the magic, you were pouring for me Embracing until we hear April thunder roll We parted beneath the cherry blossom tree Falling pink petals as I was…

According to needs for Data Science roles Python will be the number one programming language in this field for the next 5 years for sure. Due to very easy syntax and already created libraries for…