This project required that I take on the role of an entry level data analyst for the New York Citi Bike progam. My resposability was to generate a regular report for city officials that looked at user data to see where imporvements in the program could be made. I utilized Citi Bikes infrastructure for collecting data and took Citi Bike user data from 2014. I wanted to look at the difference between Summer and Winter seasons so I used the months of January, February, March, July, August and September in my report. There are currently no dashboards or sophisticated reporting processes and I have to generate some starter reports. Since my report is going to be read by city officials, public administrators and heads of New York City departments, I have to make my reports easy to read.
For the first task in this project I had to collect sets of data from the Citi Bike website. The data was not guarenteed to be consitently formatted or the size I needed so I had to do some restructuring of the data to work with it in Tableau. The steps I took to aggregate the website data to a dataset that I could use in Tableau are as follows.
- Download datasets from Citi Bike website
- Review all datasets header names and colomn arrangements to ensure they were identical
- Reduce size of each months dataset to 20,000 records for bandwidth issues
- Create Jupyter Notebook to import and combine all datasets into one document and export file
Once I had the dataset created I loaded it into Tableau to build some visualizations. Befor I built any visualizations I had to create some additional tables from the dataset to include such thing as trip time in minutes, number of trips, gender type etc. I used such things as calculations and groupings to accomplish these tasks. After completing that, I had to create at least two dashboards with a minimum of two visualizations for each Dashboard so I had to make at least four visualizations in total for this project. Although I did not end up using all of my visualizations in the report, I did end up creating 20 visualizations as part of this project. The following are the questions that I answered in the visualization creation portion of this report.
- What were the total trips taken
- What were the total trips taken by demographic age group
- What were the total trips taken by gender group
- How many of each age demographic was included in each gender group
- Did guests or subscribers use the service more
- What was the average trip durration
- What was the average trip durration by demographic age group
- Station usage by gender group
- Station usage by age demographic group
- What is the total volume usage of all start station locations
- What is the total volume usage of all end station locations
- What is the total volume usage by sesaon of start station locations by season
- What are the top ten start stations: named and mapped
- What are the top ten end stations: named and mapped
- What are the bottom ten start stations: named and mapped
- What are the bottom ten end stations: named and mapped
With the visualizations completed, I now had to create a set of Dashboards that could be used in the final Tableau Public report. With the aid of the visualizations, I needed to indentify some phenomena for each Dashboard. In addition to having a visual representation of the phenomena, I also needed to give a written explanation of what was occuring. The steps that I took to create the Dashboards are as follows.
- Create unique headers
- Drag and drop visualizations onto the dashboard
- Arrange visualizations to be coherent, easy to read and visually appealing
As part of the report, city officials have request that there be a static or dynamic map as part of the project. For the project I choose to do a static map that showed all of the bike stations with a visual representation of the most to least popular start and end locations. As part of the requirements, they asked that zip code information be overlaid on top of the map. I choose to make three static maps for this portion of the project. One showed all of the start locations data, the second showed all of the end locations data and the third showed all of the seasonal data at the start locations. There also needed to be a written explanation of what trend was observed in this visualization. The steps that I took to create these visualizations are as follows.
- Create a new worksheet
- Add start/end longitude and latitude information
- Create number of trips count marks and have them represent the size of mark based on use frequency
- Add start station, end station or seasonal table to marks section having them represent the color of the marks
At this point all of my visualizations, maps and dashboards have been created. Now I had to put all of these together into a story that could be published to Tableau Public. This is what would be presented to the officials so like all other aspects of the project, it had to be easy to read, logical and appealing. All written observations needed to be backed up by data and the report needed to be professional. The steps I took to complete this task are as follows.
- Create a new story
- Create in introduction page and summary page
- Drag created dashboards into story in sequential order of data observed
- Type up summary and add links to reference material and README.md file
Now that the report was finished I had to publish it to Tableau Public and provide a link to the city officials so that they could review it. The report had to include a minimum of
- 4-10 total visualizations that supported observed phenomenon
- Two Dashboards
- One city official map
- One story
- A text or README.md file with the analysis of the phenomenons uncovered
- Published report to Tableau Public as a .twbx file
The Citi Bike regular report will be assessed on the following metrics:
- Analytic Rigor
- Readability
- Visual Attraction
- Use any and all tools needed
- Analyze datasets thoroughly
- Use appropriate time-sets
- Build visualizations with small datasets
- Have clear answers and don't cram everything in
- Use colors and themes that match the companies aesthetic tones
- Pay attention to labels
- Keep an eye out for false data
UNIVERSITY OF OREGON: Data Analytics Boot Camp - Repository for project 16(Tableau Challenge)
Mathew Miller © 2021. All Rights Reserved.