Data Analysis on Citi Bike Trip History Through the Seasons

Posted on Apr 1, 2020
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.

Introduction

Bike-sharing programs have been around for more than 50 years and are becoming increasingly popular. However, when weather strikes, it can take a toll on the bikes and the people that ride them. Therefore, in an effort to examine the impact of seasonality on bike-share ridership, I gathered an analyzed data from Citi Bike in New York, NY.

The data consisted of 79 files containing more than 80 million records of user trip histories from the launch of Citi Bike in 2013 through 2019. Each file represented the monthly trip history for a given month and year. Since the data was quite large, I wrote a function to sample 2% of the data from each file and concatenated the sampled data together. In conclusion, I was left with approximately 1.83 million observations from June 2013 (launch) through 2019. This method was chosen so an even spread of observations could be obtained across time.

First, I will discuss a brief history of bike-sharing programs and Citi Bike. Then I will explore the data and characteristics of Citi Bike users. Finally, I argue that there is a seasonal trend that impacts the number of trips. 

 

Background

Demaio (2009) noted that three major shifts in bike-sharing programs. The first program was observed in the Netherlands in the 1960s. This was orchestrated by a Dutch fringe group known as Provo in an effort to provoke the establishment and ban cars from the city of Amsterdam. The plan was to gather old bikes, paint them white, and leave them throughout the city. The group provided 50 bikes that were to always remain unlocked and available for public use. However, complications ensued and the plan fell apart in a matter of days. 

The second shift occurred in Denmark in the early to mid 1990s according to Demaio (2009). Several cities started their own bike-sharing programs and one of the technologies utilized was the development of docking stations. Additionally, bikes started becoming more durable and they featured advertising space. More durable bikes allowed for greater use. Furthermore, partnerships could be developed or increased revenue could be generated by having the ability to sell advertising space. Finally, technological advancements inspired the third shift in bike-sharing programs. Users were able to utilize magnetic strip cards to rent bikes and docking stations had electronic locks. These transformations led to the bike-sharing programs observed today.

 

Citi Bike

Enter Citi Bike. Citi Bike (2020a) is the bike-sharing program in New York, NY and the program was launched in May 2013. Citi Bike is the nation’s largest bike-sharing program consisting of approximately 13,000 bikes and 850 docking stations. The program operates all year in Manhattan, Brooklyn, Queens, and Jersey City. It is important to note that this analysis did not include data from Jersey City. Citi Bike (2020b) provides several options for people to participate in the bike-sharing program. One option is to purchase an annual membership that covers an unlimited number of rides 45-minutes or less. Less frequent users or tourists, have options that include a single trip, a single day pass, or a 3-day pass.

 

Data Analysis

User Data Analysis

Demographic information is fairly limited in the data, however, approximately 67.5% of riders identified as male, 22.3% as female, and 10.2% as unknown. Additionally, about 88% of riders were annual subscribers. Thus, there are more male riders and more annual subscribers.

Cycling through the Seasons: Data Analysis on Citi Bike Trip History Cycling through the Seasons: Data Analysis on Citi Bike Trip History

For all users, the peak trip time began between 7 and 9 a.m. and trips ended between 4 and 6 p.m. indicating that people may use this as a transportation method before and after regular working hours. A similar pattern was observed for users that identified as male or female.

Cycling through the Seasons: Data Analysis on Citi Bike Trip History

Additionally, the same peak times were observed in annual subscribers, however, there was never a peak for non-subscribers. This aligns with the majority of users being annual subscribers and likely utilizing bikes for commuting to work. Since non-subscribers are not likely utilizing Citi bikes on a regular basis it makes sense to not see a peak in the frequency of use by hour for this group. 

Data on Frequency of Trips

Additionally, the figure below displays the frequency of trips per day and per month by each user type. As mentioned previously, subscribers have far more trips. Furthermore, a potential seasonal effect can be observed in both user types. 

In addition to the starting hour of trips, I also examined the frequency of trips by day of the week. Similar to peak times, most trips occurred during weekdays compared to weekends with the highest frequency of trips occurring on Wednesday.

To examine day of the week further, days were categorized into weekday or weekend and then explored by the hour of the starting trip. Similar trends were observed in the starting hour. Weekday trips peaked between 7 and 9 a.m. and trips ended between 4 and 6 p.m. Alternatively, there was no major peak observed in weekend trips. A clear picture is provided when considering each of these factors together: peak times are likely people commuting to and from work during the week.

 

Data on Seasonality

In order to get a better sense of the impact of seasonality on the frequency of bike trips, I gathered weather data from NOAA for observations taken at Central Park. The data consisted of daily summaries including wind, precipitation, and the minimum and maximum daily temperatures. Below is a plot that contains the daily minimum and maximum temperatures. There is a clear pattern of decreasing temperatures in the winter months as would be expected.

Furthermore utilizing simple linear regression, I examined the impact of temperature and precipitation on the frequency of bike trips. There is a clear linear relationship between an increase in temperature and an increase in the frequency of trips. Furthermore, as precipitation increases the frequency of trips also decreases. However, there are a few outliers on days with larger amounts of precipitation that could be impacting the regression.

The following plot shows the frequency of bike trips each day. There is a similar pattern of decreasing trips in the winter months. Noticeably, it appears the frequency of trips is generally increasing over time. The trend in the frequency of trips was calculated utilizing a rolling average method. The trend line makes the increases in the frequency of trips even more clear. Even though there are decreases in the frequency of trips in the winter months Citi Bike has been growing its ridership overall.

Additionally, with the trend line calculated, it is possible to decompose the seasonal effects. By removing the trend from the observed data greater insight can be derived by confirming a seasonal effect on the frequency of bike trips. The plot below aligns well with the observed data, however, the seasonal effect is even more noticeable. Thus, there is a clear indication that seasonality does impact the frequency of bike trips.

 

Conclusion

 Although bike-sharing programs have been in the works for more than 50 years, they are becoming increasingly popular as ways to bridge the gaps left by other methods of transportation. As indicated above, there is a clear seasonal trend the impacts the frequency of trips. Even with the seasonal effects of the weather, Citi Bike has been able to increase its ridership since its launch in 2013. In order to maintain this trend it is important for Citi Bike to not only consider the seasonal impact on its equipment but also its users.

 

Future Work

Future work to examine seasonality could continue. Alternative models of seasonal decomposition could be done to check for greater accuracy. Additionally, non-linear models could be used to predict the frequency of trips over time. Non-linear models are important given the daily ridership data is not linear as observed in the seasonal effects and overall ridership data. Furthermore, a more granular investigation could done to include hourly precipitation and snowfall. This may provide a better picture of ridership at a given time.

 

References

Citi Bike. (2020a). About Citi bike. Retrieved from https://www.citibikenyc.com/about.

Citi Bike. (2020b).  Pricing. Retrieved from https://www.citibikenyc.com/pricing

Demaio, P. (2009). Bike-sharing: History, impacts models of provision, and future. Journal of Public Transportation. 12(4).

About Author

Tyler Kotnour

Tyler Kotnour graduated with his Master of Public Administration degree in 2018. Upon graduation, Mr. Kotnour worked in consulting conducting research and program evaluation. His primary role involved analyzing public health data for governments and non-profits. A major...
View all posts by Tyler Kotnour >

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI