The flights data were downloaded from BTS - U.S. Department of Transportation (US DOT), Bureau of Transportation Statistics. The file for each year was renamed to BTS<year> ( BTS1990.csv, BTS1991.csv, …, BTS2014.csv ) and collected in a subdirectory BST.
The row structure in data tables is the following:
| YEAR | Year |
| QUARTER | Quarter |
| MONTH | Month |
| ORIGIN_AIRPORT_ID | Origin Airport, Airport ID. An identification number assigned by US DOT to identify a unique airport. Use this field for airport analysis across a range of years because an airport can change its airport code and airport codes can be reused. |
| ORIGIN_AIRPORT_SEQ_ID | Origin Airport, Airport Sequence ID. An identification number assigned by US DOT to identify a unique airport at a given point of time. Airport attributes, such as airport name or coordinates, may change over time. |
| ORIGIN_CITY_MARKET_ID | Origin Airport, City Market ID. City Market ID is an identification number assigned by US DOT to identify a city market. Use this field to consolidate airports serving the same city market. |
| ORIGIN | Origin Airport |
| ORIGIN_CITY_NAME | Origin Airport, City Name |
| ORIGIN_COUNTRY | Origin Airport, Country |
| ORIGIN_COUNTRY_NAME | Origin Airport, Country Name |
| ORIGIN_WAC | Origin Airport, World Area Code |
| DEST_AIRPORT_ID | Destination Airport, Airport ID. An identification number assigned by US DOT to identify a unique airport. Use this field for airport analysis across a range of years because an airport can change its airport code and airport codes can be reused. |
| DEST_AIRPORT_SEQ_ID | Destination Airport, Airport Sequence ID. An identification number assigned by US DOT to identify a unique airport at a given point of time. Airport attributes, such as airport name or coordinates, may change over time. |
| DEST_CITY_MARKET_ID | Destination Airport, City Market ID. City Market ID is an identification number assigned by US DOT to identify a city market. Use this field to consolidate airports serving the same city market. |
| DEST | Destination Airport |
| DEST_CITY_NAME | Dest Airport, City Name |
| DEST_COUNTRY | Destination Airport, Country |
| DEST_COUNTRY_NAME | Destination Airport, Country Name |
| DEST_WAC | Destination Airport, World Area Code |
| UNIQUE_CARRIER | Unique Carrier Code. When the same code has been used by multiple carriers, a numeric suffix is used for earlier users, for example, PA, PA(1), PA(2). Use this field for analysis across a range of years. |
| AIRLINE_ID | An identification number assigned by US DOT to identify a unique airline (carrier). A unique airline (carrier) is defined as one holding and reporting under the same DOT certificate regardless of its Code, Name, or holding company/corporation. |
| UNIQUE_CARRIER_NAME | Unique Carrier Name. When the same name has been used by multiple carriers, a numeric suffix is used for earlier users, for example, Air Caribbean, Air Caribbean (1). |
| UNIQUE_CARRIER_ENTITY | Unique Entity for a Carrier's Operation Region. |
| REGION | Carrier's Operation Region. Carriers Report Data by Operation Region. |
| CARRIER | Code assigned by IATA and commonly used to identify a carrier. As the same code may have been assigned to different carriers over time, the code is not always unique. For analysis, use the Unique Carrier Code. |
| CARRIER_NAME | Carrier Name |
| CARRIER_GROUP | Carrier Group Code. Used in Legacy Analysis |
| CARRIER_GROUP_NEW | Carrier Group New |
| DISTANCE_GROUP | Distance Intervals, every 500 Miles, for Flight Segment |
| CLASS | Service Class |
| PASSENGERS | On-Flight Market Passengers Enplaned |
| FREIGHT | On-Flight Market Freight Enplaned (pounds) |
| On-Flight Market Mail Enplaned (pounds) | |
| DISTANCE | Distance between airports (miles) |
0 "PASSENGERS" 0.00 3167.00 66206.00 1 "FREIGHT" 118745.00 428.00 3106805.00 2 "MAIL" 0.00 0.00 0.00 3 "DISTANCE" 4513.00 1037.00 3451.00 4 "UNIQUE_CARRIER" "ADB" "YV" "BA" 5 "AIRLINE_ID" 20110 20378 19540 6 "UNIQUE_CARRIER_NAME" "Antonov Company" "Mesa Airlines Inc." "British Airways Plc" 7 "UNIQUE_CARRIER_ENTITY" "9488B" "11022" "9493E" 8 "REGION" "I" "L" "I" 9 "CARRIER" "ADB" "YV" "BA" 10 "CARRIER_NAME" "Antonov Company" "Mesa Airlines Inc." "British Airways Plc" 11 "CARRIER_GROUP" 0 2 0 12 "CARRIER_GROUP_NEW" 0 2 0 13 "ORIGIN_AIRPORT_ID" 11548 14107 12478 14 "ORIGIN_AIRPORT_SEQ_ID" 1154804 1410702 1247802 15 "ORIGIN_CITY_MARKET_ID" 31548 30466 31703 16 "ORIGIN" "EMA" "PHX" "JFK" 17 "ORIGIN_CITY_NAME" "Derby, United Kingdom" "Phoenix, AZ" "New York, NY" 18 "ORIGIN_COUNTRY" "GB" "US" "US" 19 "ORIGIN_COUNTRY_NAME" "United Kingdom" "United States" "United States" 20 "ORIGIN_WAC" 493 81 22 21 "DEST_AIRPORT_ID" 15530 11874 12972 22 "DEST_AIRPORT_SEQ_ID" 1553002 1187404 1297203 23 "DEST_CITY_MARKET_ID" 35530 31874 30730 24 "DEST" "UXP" "GDL" "LHR" 25 "DEST_CITY_NAME" "Bay St. Louis, MS" "Guadalajara, Mexico" "London, United Kingdom" 26 "DEST_COUNTRY" "US" "MX" "GB" 27 "DEST_COUNTRY_NAME" "United States" "Mexico" "United Kingdom" 28 "DEST_WAC" 53 148 493 29 "YEAR" 2014 2014 2014 30 "QUARTER" 1 2 2 31 "MONTH" 1 5 5 31 "DISTANCE_GROUP" 10 3 7 32 "CLASS", "P" "F" "F"
# flightsNet.py
# transforming the monthly flights data 1990-2014 from
# http://www.transtats.bts.gov/DL_SelectFields.asp?Table_ID=260
# into Pajek's multirelational temporal file
# by Vladimir Batagelj, January 22, 2015
import csv, sys, os, datetime
global air, airports, cmp, companies
def indAirport(name):
global air, airports
if not(name in air.keys()):
air[name] = len(air)+1
airports.write(str(len(air))+' "'+name+'"\n')
return air[name]
def indCompany(name):
global cmp, companies
if not(name in cmp.keys()):
cmp[name] = len(cmp)+1
companies.write(str(len(cmp))+' "'+name+'"\n')
return cmp[name]
print("MakeAirTen *********")
print('start:',datetime.datetime.now().ctime())
workdir = 'E:/data/airports/'
os.chdir(workdir)
arcs = open('arcs.tmp', 'w')
arcs.write('*arcs \n')
airports = open('airports.tmp', 'w')
airports.write('*vertices \n')
companies = open('companies.tmp', 'w')
companies.write('*vertices \n')
cmp = {}; air = {}
line = 0
for year in range(1990,2015):
podatki = './BTS/BTS'+str(year)+'.csv'
with open(podatki, newline='', encoding='utf-8') as dat:
airReader = csv.reader(dat, delimiter=',', quotechar='"')
try:
head = next(airReader)
for row in airReader:
arcs.write(str(indCompany(row[4]))+': '+str(indAirport(row[16]))+ \
' '+str(indAirport(row[24]))+' '+str(row[0])+' ['+ \
str(12*(int(row[29])-1990)+int(row[31]))+']\n')
line = line+1
except csv.Error as e:
sys.exit('file {}, line {}: {}'.format(podatki, airReader.line_num, e))
print(year,':',datetime.datetime.now().ctime())
print("lines = ",line,"\nairports = ",len(air),"\ncompanies = ",len(cmp))
arcs.close(); airports.close(); companies.close()
print('stop:',datetime.datetime.now().ctime())
>>> MakeAirTen ********* start: Thu Jan 22 04:45:14 2015 1990 : Thu Jan 22 04:45:15 2015 1991 : Thu Jan 22 04:45:17 2015 1992 : Thu Jan 22 04:45:21 2015 1993 : Thu Jan 22 04:45:25 2015 1994 : Thu Jan 22 04:45:27 2015 1995 : Thu Jan 22 04:45:28 2015 1996 : Thu Jan 22 04:45:30 2015 1997 : Thu Jan 22 04:45:32 2015 1998 : Thu Jan 22 04:45:33 2015 1999 : Thu Jan 22 04:45:35 2015 2000 : Thu Jan 22 04:45:37 2015 2001 : Thu Jan 22 04:45:39 2015 2002 : Thu Jan 22 04:45:43 2015 2003 : Thu Jan 22 04:45:45 2015 2004 : Thu Jan 22 04:45:47 2015 2005 : Thu Jan 22 04:45:49 2015 2006 : Thu Jan 22 04:45:51 2015 2007 : Thu Jan 22 04:45:55 2015 2008 : Thu Jan 22 04:45:57 2015 2009 : Thu Jan 22 04:45:59 2015 2010 : Thu Jan 22 04:46:02 2015 2011 : Thu Jan 22 04:46:04 2015 2012 : Thu Jan 22 04:46:06 2015 2013 : Thu Jan 22 04:46:08 2015 2014 : Thu Jan 22 04:46:08 2015 lines = 1737054 airports = 2776 companies = 633 stop: Thu Jan 22 04:46:11 2015
# transforms a Pajek's temporal network in temporal matrix
# Vladimir Batagelj, 22. Jan 2015
import os, re, pickle, datetime
from TQ import *
os.chdir('E:/data/airports/')
print('start:',datetime.datetime.now().ctime())
net = open('flights.net','r')
n = 2776
ten = open('flights.ten', "w")
ten.write("*nodes "+str(n)+"\n")
while True:
line = net.readline()
if line[0]=='*': break
k = 0; names = ['']*n; C = [[(1,296,1)]]*n
while True:
line = net.readline()
if (line[0]=='*') or (k>=n): break
L = re.split('\s+',line.strip())
j = eval(L[0])-1; names[j] = L[1][1:-1]
ten.write(L[0]+' '+L[1]+' [(1,296,1)]\n')
k = k+1
while True:
line = net.readline()
if line[0]!='*': break
print('nodes done:',datetime.datetime.now().ctime())
W = [[ [] for u in range(n)] for v in range(n)]
ten.write("*arcs\n")
while line:
L = re.split('\s+',line.strip())
u = eval(L[1])-1; v = eval(L[2])-1
# if v<u: (u,v) = (v,u)
w = eval(L[3]); t = eval(L[4][1:-1])
W[u][v] = TQ.sum(W[u][v],[(t,t+1,w)])
line = net.readline()
for u in range(n):
for v in range(n):
if W[u][v]!=[]:
ten.write(str(u+1)+' '+str(v+1)+' '+str(W[u][v])+'\n')
net.close()
ten.close()
print('arcs done:',datetime.datetime.now().ctime())
Months = ['Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec' ]
time = [ v+str(u)[2:] for u in range(1990,2015) for v in Months ][:294]
flights = {}
flights['tit'] = 'Flights'
flights['num'] = n
flights['nam'] = names
flights['act'] = C
flights['tim'] = time
flights['mat'] = W
with open("flights.p","wb") as p: pickle.dump(flights,p)
print('saved:',datetime.datetime.now().ctime())
start: Thu Jan 22 23:47:10 2015 nodes done: Thu Jan 22 23:47:10 2015 arcs done: Thu Jan 22 23:53:11 2015 saved: Thu Jan 22 23:53:15 2015
The airports' attribute data can be found at Open flights site or datahub.
To list of airports in flights.ten available in airports.tmp we assign properties (attributes) from airportsHead.csv . We also include the numeric code from countries.dat for countries:
# constructs airports' attributes
# vladimir Batagelj, January 24, 2015
import os, re, csv, datetime
global air, cnt
def indAirport(name):
global air
if not(name in air.keys()): air[name] = len(air)+1
return air[name]
def indCountry(name):
global cnt
if not(name in cnt.keys()): cnt[name] = len(cnt)+1
return cnt[name]
os.chdir('E:/data/airports')
print('start Construct:',datetime.datetime.now().ctime())
print('Countries')
cnt = {}; air = {}
countries = open('countries.dat', 'r')
row = countries.readlines()
countries.close()
for r in row: ind = indCountry(r[:-1])
line = 0
with open('airportsHead.csv', newline='', encoding='utf-8') as dat:
airReader = csv.reader(dat, delimiter=',', quotechar='"')
try:
head = next(airReader)
print('|'.join(head))
for row in airReader:
nam = row[1]; cny = row[3]
iata = row[4]; icao = row[5]; tiz = row[9]
lat = row[6]; lon = row[7]; alt = row[8]
icny = indCountry(cny)
if len(iata)>2:
air[iata] = ( nam, iata, icao, icny, cny, lat, lon, alt)
elif len(icao)==3:
air[icao] = ( nam, iata, icao, icny, cny, lat, lon, alt)
else: print('missing data', line, nam)
line = line+1
except csv.Error as e:
sys.exit('file {}, line {}: {}'.format(podatki, airReader.line_num, e))
print("lines = ",line,"\nairports = ",len(air))
attrs = open('airports.atr', 'w')
attrs.write('*vertices \n')
airports = open('airports.tmp', 'r')
rows = airports.readlines()[1:]
airports.close()
for r in rows:
s = r.split(' ')
ap = s[1][1:-2]; an = int(s[0])
if ap in air.keys():
( nam, iata, icao, icny, cny, lat, lon, alt) = air[ap]
attrs.write(s[0]+';"'+nam+'";"'+iata+'";"'+icao+'";'+str(icny)+';"'+ \
cny+'";'+str(lat)+';'+str(lon)+';'+str(alt)+'\n')
else:
print("not found", an, ap)
attrs.write(s[0]+';"?";"'+ap+'";"'+ap+'";0;"?";0;0;0\n')
attrs.close()
print('finished:',datetime.datetime.now().ctime())
import os, re, datetime
from TQ import *
from operator import itemgetter
os.chdir('E:/data/airports')
print('start Extract:',datetime.datetime.now().ctime())
print('Network Flights')
AIR = TQ.Ianus2Mat("./Flights.ten")
print('input done:',datetime.datetime.now().ctime())
mat = AIR['mat']
print('sum =',TQ.MatSummary(mat))
print(AIR.keys())
print('dim =',AIR['dim'])
print('met =',AIR['met'])
print('nam =',AIR['nam'][0:10])
print('tit =',AIR['tit'])
print('typ =',AIR['typ'])
print('tin =',AIR['tin'][0:10])
print('til =',AIR['til'][0:10])
(nr,nc,tmin,tmax) = AIR['dim']
n = len(mat); All = range(n)
names = AIR['nam']
A = [[]]*n; S = [0]*n
for u in range(n):
a = []
for v in range(n):
a = TQ.sum(a,mat[v][u])
A[u] = a; S[u] = TQ.total(a)
P = sorted(enumerate(S),key=itemgetter(1),reverse=True)
for i in range(100):
print(i+1,names[P[i][0]],P[i][1])
B = [[ [] for u in range(100)] for v in range(100)]
L = []; nam = []
for (i,v) in P[0:100]:
L.append(i); nam.append(names[i])
for (i,u) in enumerate(L):
for (j,v) in enumerate(L):
B[i][j] = mat[u][v]
best = {}
best['mat'] = B
best['dim'] = (100,100,1,295)
best['met'] = AIR['met']
best['nam'] = nam
best['tit'] = "100 largest airports"
best['typ'] = AIR['typ']
best['til'] = AIR['til'][0:100]
best['tin'] = AIR['tin']
TQ.MatSave(best,file="best100.ten")
print('finished:',datetime.datetime.now().ctime())