A Data Foundation Nanodegree Course at Udacity.
For my final project during this introductory course into Data Analytics and Data Visualisation, I analysed a set of US Aviation data to answer the following questions.
Question 1:
How did airlines vary on the busiest flight path in the US in 2015 in terms of Delays and Day of Week?
I have created a story in Tableau that will find an answer to this question in 3 steps:
Slide 1: Which was the busiest flight path in the US in 2015?
Slide 2: Which airlines operated on this connection and how many flights did they offer?
Slide 3: Adding all delays together, which airline accumulated the longest delays and on which day of the week?
Main insights:
Slide 1: The connection with the most flights operated in the US in 2015 was from San Francisco to Los Angeles.
Slide 2: These 6 airlines had flights on this route and with 201 flights United Air Lines offered the most in 2015.
Slide 3: By far, Delta Air Lines accumulated the longest delays (with an average Total Travel Time of more than 160mins) and operated particularly bad on Sundays, Wednesdays and Saturdays.
Design:
For this visualisation I have chosen the stacked bar chart, because I wanted to show how delays are literately getting “stacked up” during air traffic and to compare which airlines are most affected.
Question 2:
Which airline offers the best service throughout the year between Anchorage and Seattle?
Main insight:
Further insights:
- JetBlue Airways offer the smallest number of flights of just 7-9 in both directions and only operate between May and September. The longest Avg. Arrival Delay has been almost 30mins.
- Same as Alaska Airlines, Delta Air Lines offer flights throughout the year with a surprising avg. performance in January (arriving 40mins early, Anchorage -> Seattle), but struggle with long delays in March and November (Seattle -> Anchorage).
Design:
In Line and Area Charts the audience can easily compare a year month-by-month and spot irregularities quickly. To visualize an overall total, like the sum of flights being offered each month, the Area Chart comes in particularly handy.
Question 3:
How is the relationship between Departure Delays and Arrival Delays across all American airlines on average and how is it affected by the Time of Year (Month)?
Main insights:
- Looking at 2015 in total, it seems that most airlines have clustered in the center of the chart, within a Departure Delay range of 8-12 mins and an Arrival Delay range of 4-7 mins.
- Looking at each month individually though, the avg. delay times vary quite significantly.
- It appears that June and December are prone to longer delays than other months, while in September all airlines perform much better than the rest of the year.
Further insights:
- Some “smaller” airlines appear to perform very well, like Hawaiian and Alaska Airlines, while other ones, like Spirit and Frontier Airlines, seem to struggle with delays all the time.
- For some reason, US Airways only operated in the first half of 2015, from January till June.
Design:
As we all know, size is usually not a very good encoding for data. In this Scatter Chart, I only used spheres of different sizes as a rough indicator of how many flights the different airlines offer and which ones can be considered as “bigger” or “smaller” airlines. For an accurate comparison of the number of flights, I have added a Bar Chart at the bottom. The Scatter Chart though allows us to easily spot clusters, trends and outliers in the dataset, especially over time with the provided monthly filter.
