Final Assignment – London Underground Traffic

on 02|02|2015
Filled under: Uncategorized

 

The main dataset used was the same as the previous assignment, which were the tickets scans for each station of the London Underground per fifteen minute period. For this assignment there is a slider that can be used to show values across the entire day. This form of visual aid was chosen over animation as it gives more interactivity to the user.

Overall1530 example

 

The above comparison shows the how the slider can be used to quickly identify the changes over the course of a day. A rollover function is also present to identify the exact value of the station. As expected the busiest times in the system is at around 08:00 – 09:30 in the morning and 17:00-19:00 in the evening. Other trends spotted was that many stations around the city centre (eg Oxford St. and Piccadilly Circus) had the most traffic during the evening indicating that there was large amounts of nightlife in that area.

Another strange data point was Canary Wharf. During morning rush hour there was relatively small amounts of traffic but during the evening there was a large spike. This was unexpected as Canary Wharf is predominantly offices and would be expected to have a ‘nine to five’ workday. This could lead to more analysis into other transport systems and find out what is happening at that data point.

0900 example

A limitation of the data was that although you could see the amount of people in each station, it was difficult to interpret which direction the travel was occurring. It was intended that the visualisation could toggle between the different datasets, however issues with reading the different arrays made this difficult for me to code. With this information it would have shown the direction of flow through the system.

max, avg toggle

 

To analyse the data over the course of a day, the visualisation used the maximum and averages of each station which was calculated in the dataset.  This value again could be found using the rollover function. This allowed for a value at a certain time to be viewed then toggled, comparing it to the daily average or daily maximum. There was also a total counter, which varied over time, showing the total number of users in the system. The only issue was that the code printed all the values of each station over each over meaning it was illegible. This error was unable to be corrected.

This assignment was able to show much more information compared to the previous assignment of which the code was built on. It clearly shows the traffic through the London Underground across a typical weekday. It could be expanded to show weekend values and also show in or out ticket scans as described previously. However I believe the data and trends can be quickly identified due to the clarity of the visualisation.

To further improve this dataset, it would be beneficial to evaluate other modes of transport. The London Datastore has a sample of Oyster Card uses that include a start and end point. This could be used to show the migration of multiple transport types across London. The visualisation could include an animation option that shows people (represented as dots) travelling across the city. This could fill some gaps in the trends that have been spotted, such as at Canary Wharf, and bring a deeper analysis to the visualisation.

 

Comments are closed.