Drugs
Routes
Data comes from UNODC, apparently based on law enforcement data on individual arrests between 2011-2016. If they caught the smuggling en route from-to-country they are usually recorded. Sometimes source, target is missing, but if they recorded the production country, code takes that as source, if destination is missing, country of arrest is used.
Report treats the weights of all drugs as equivalent, they are summed per route, to give a general idea of the importance of each route. Code also multiplies tablet based drugs with 100 mg per tablet, to turn them into a kilo weight. Cannabis arrests are removed to focus on harder drugs.
Routes are drawn from country's center coordinate to another's center coordinate, so the places seen at end points of routes are nothing special (not cities, nor ports).
A quick sum on transport method during arrest;
import zipfile, pandas as pd
with zipfile.ZipFile('drug-trafficking-unodc.zip', 'r') as z:
df = pd.read_csv(z.open('drug-trafficking-unodc.csv'),sep=';')
print (df.groupby('TRANSPORT').size().sort_values(ascending=False))
TRANSPORT
Commercial air 19137
Land 8933
Private road 6933
Other 5942
Commercial road 3471
Unknown 2833
Commercial sea 1021
Mail 924
None 534
Rail 368
Private air 57
River 30
Private sea 24
Pedestrian 16
Air 6
dtype: int64