Information for data science project

[fabianosilva]

Hello there,

If this is not right place I ask the moderators to move this thread or delete it.

I'm doing a data science project trying to decompose the price structure of air tickets.
https://rpubs.com/fabiano_silva/RoundTheWorld
I'm using data from the US Bureau of transport and statistics

My intention is to prove that the longer the flight, cheap it is (relative to distance). During the process I found that most tickets keep a range from the cheaper to the most expensive at exactly 20 times, no matter airline company.
You can see it here:

View attachment 23279

The graph is distance (Miles) on X and price per km flown on Y (sorry for mixed units!), both on log10 scales.

I need some opinion from people inside the industry or logistics on why they choose this range for price strategy.
Unfortunately, I don't have the temporal data so I can relate this to time of purchase or such.

Does the lower price is always the pure "cost" and then the airline raise the price according to seat number in a negative exponential?
Any one have any idea? Opinions are welcome.

Regards, Fabiano

 

[Aviador]

We don't normally allow these kind of requests but it's an interesting topic I'd be interested in learning more about if you'd care to share some of your findings when you've collated the information.

 

[fabianosilva]

We don't normally allow these kind of requests but it's an interesting topic I'd be interested in learning more about if you'd care to share some of your findings when you've collated the information.

Thanks!

Until now, someone who works researching the Air Travel industry refused my theory on pure cost on lower values.

It's interesting to think that on long flights (+4000 miles), you could pay 0.06 USD/Mile and they still have profit!

I'll update here with any conclusions.