This repository contains an additional synthetic testing dataset (based on realistic characteristics of data from a major carrier in 2016) for the paper "Choice-Based Airline Schedule Design and Fleet Assignment: A Decomposition Approach" published on Transportation Science. If you use any of the material here, please include a reference to the paper and this webpage.
Flights | Itineraries | Fare Products | Markets | Fleet Types | Aircraft | Market Share |
---|---|---|---|---|---|---|
815 | 4,290 | 47,190 | 819 | 7 | 186 | 42.36% |
-
flight.json
key: flight ID
value: {"origin": origin airport; "destination": destination airport; "depttime": departure time in hhmm format; "arrtime": arrival time in hhmm format}
-
product.json
key: product ID
value: {"fare": price of the fare product; "cabin": cabin of the fare product, {'Y': economy, 'C': business, 'F': first}; "origin": origin airport; "destination": destination airport; "demand": unconstrained demand/attractiveness value; "market": origin-destination market; "leg": list of flight IDs which this fare product uses}
-
market.json
key: market ID, origin+destination
value: {"total_demand": total demand/attractiveness value in this origin-destination market (including competing airline); "OA_demand": competing airlines' demand/attractiveness value in this origin-destination market}
-
fleet.json
key: fleet ID
value: {"FCAP": seats in F cabin; "CCAP": seats in C cabin; "YCAP": seats in Y cabin; "hourly_cost": hourly operating cost; "availability": number of aircraft available for this fleet type}
-
Minimum Turn Time: 35 minutes
-
Subnetwork partition (, ), see paper for notation (cf. Table 3).
-
5 hour CPU time, 2.3 GHz Quad-Core Intel Core i7, Gurobi 9.02, see paper for notation (cf. Table 4).