Udacity_Data-Analyst-Course-project-3

I worked on USA Airline Data for 2008 from the Harvard Dataverse. The dataset can be found Here

Dataset

The data consists of flight arrival and departure details for all commercial flights within the USA in the year 2008. It consist of 2389217 rows and 29 columns. There are 16 columns (CancellationCode, CarrierDelay, WeatherDelay, NASDelay, SecurityDelay, LateAircraftDelay, ArrTime, ActualElapsedTime, CRSElapsedTime, AirTime, ArrDelay, TaxiIn, TaxiOut) with missing data, but i dropped only 6 columns. This is because the values with missing data i dropped is of no value in my analysis while they rest columns i didn't drop is of value in my analysis and has less missing data. Also, if i had dropped all the features with missing data it will disrupt my analysis and i will loose important features. For eg, Diverted columns has only two distinct values with no missing data. If i had dropped the columns with missing values, i will loose one of the distinct values in Diverted column. Here are the questions i want to find answers to in my data explanatory analysis:

• Are there certain destination or arrival cities that are home to more delays or cancellations?
• What are the preferred times for flights to occur? Are there any changes over multiple years?

Summary of Findings

For this analysis, I focused exclusively on the 2008 airline dataset due to the file’s size and the limitations of my local machine’s processing capacity. The dataset consists of 1,804,634 rows and 25 columns, and I concentrated on key features including: Destination (Dest), Time Taken, Cancellation Code, Departure Delay (DepDelay), Origin, Cancelled, and Month.

My analysis explored the top 10 origin and destination cities, the duration of flights (Time Taken), monthly travel trends, and average delay times per carrier. One of the key insights revealed that the most efficient months for air travel were March and April, with average flight durations of 161.23 minutes and 161.82 minutes respectively, indicating these months are optimal in terms of flight time to destination.

Key Insights for Presentation

I analyzed the top 10 origin and destination cities and found that while none of them recorded any cancellations, several experienced notable delays. Among the destination cities, ORD (Chicago O'Hare) and DTW (Detroit Metropolitan) had the highest instances of delays. On the origin side, LAS (Las Vegas) and PHX (Phoenix) showed the most delays. I also examined the months in which these trips occurred to identify any seasonal patterns influencing the delays.