Nice Ride Minnesota is a bikesharing service with 3000+ bikes and 400+stations across Minneapolis. These bikes are faster than other ways of getting around. They can be conveniently dropped at any nearby Nice Ride station at the end of the ride. A single ride costs $2, a day’s pass with unlimited 30-minutes rides in a 24-hour period costs $6 and annual membership costs $75.
I decided to explore the bike history data from 2017 season to see if data shows any interesting trends or patterns. The dataset for 2017 is freely available on the following link.
Nice Ride Minnesota 2017
Bike sharing trip and station data from Nice Ride MN for the 2017 year
I also found the weather dataset on the above link which made my job easier as I wanted to analyse how weather changes affect ridership. I used Tableau Public which is a data visualization software, available for free here.
Download Tableau Public
Build mobile-friendly dashboards in seconds with automatic phone and tablet layouts. Use the layouts as they come, or…
Understanding the bandwidth of the data
The Nice ride dataset contains approximately 460,00 records.
It consists of just a few fields
- Start Date — The date of starting the ride
- Start station — The station where the ride initiated
- Start station number — The start station serial number
- End Date — The date of ending the ride
- End Station — The station where the ride terminated
- End Station number — The end station serial number
- Account type — Indicates whether the user is a “member” or “casual” rider
- Total duration (seconds) — The total time the ride took place for in seconds
The total duration in seconds can be converted to minutes using calculated field in Tableau like this
Linking the Files and changing data types
I first connected the ‘Nice ride trip history 2017' database in the Tableau workspace and then joined Weather data to it. Since, date field was common among both the databases, I decided to join them using date as the ‘key’. There are two types of dates — ‘start date’ and ‘end date’ available in trip history database; we don’t really need both of them. So, I’ve considered only ‘end date’ relevant to the analysis. The ‘end date’ was in the ‘datetime’ format where as the ‘date’ field in Weather data was in data format. Tableau throws error if two fields of different data types are joined together. So I changed the end date field format from ‘datetime’ to ‘date’ and joined both the databases on the key ‘date’.
Here is a screenshot of how the joined tables look like
Data Visualization and mining trends
I explored several trends in the data such as how maximum temperature affected the number of rides, what day of the week we see maximum rides, what duration do the casual rider and commuters ride, so on and so forth. I combined all the analyses into a dashboard which can be viewed here.
- Rides and Precipitation
2. Rides and Temperature
3. Rides per Day
4. Membership rides
5. Popular start stations
6. Popular destinations
7. Duration of Rides and members
Now lets find out what story these visualizations tell us. The first two dual axis charts show that higher the precipitation (rainy days), lower is the ridership on those days. Also, the maximum temperature affects the number of rides on the days the temperature shoots through the ceiling.
I was guessing that one of the days over the weekend would be when most rides would happen. And I’m not surprised to find out that Saturday happens to be that day!
Among the number of rides made by the casual riders and the members during the 2017 season, it seems that members made the most number of rides.
Around 11,000 rides seem to have originated at Lake Street & Knox Ave S and 12,000 rides seem to have also terminated there; making it the most popular ride station. My interpretation is that most of the members who ride the Nice Ride live in that area and therefore, most of the rides are commuter rides to and fro from home to work. In the University of Minnesota area, locations such as Coffman Memorial, McNamara Center, Social Sciences and Willey Hall seem popular.
As far as the average duration of rides made by casual riders and members is concerned, data shows that casual riders tend to ride for longer duration. My guess is that casual riders tend to spend a lot of time exploring places which therefore, results in longer duration rides.
In this article, we found some interesting insights about Nice Rides ridership in Minneapolis using Tableau and by combining two publicly available datasets. Tableau made the data analysis extremely fast and the results were pretty insightful.
Thanks for reading!