An enhanced UN/LOCODE dataset with significant improvements:
The main reason this project exists: coordinates in the original UN/LOCODE list have major problems:
Only 80% of locations have coordinates
UN/LOCODEs without coordinates don't just include tiny villages, but world's most important cities like Oakland (USOAK), Barcelona (ESBCN) and London (GBLON).
Many coordinates are just wrong
Quite a few coordinates have typos (BOSCS), but many are just flat out wrong (EGSCN)
Many coordinates are automatically improved by combining the data with data from OpenStreetMap's Nominatim API and Wikidata. This is then doublechecked against multiple alternative datasources and manually corrected if needed, resulting in an exceedingly accurate dataset.
This is solved with code-list-improved.csv. It has much more accurate coordinates, and it has way more of them (98.7%; this includes all ports in World Port Index).
Next to the improved UN/LOCODEs, there are other nifty things in this repo:
An example: DEBHQ (Bahrenfeld), is in Hamburg (DEHAM), but how would you know these are essentially the same place?
For this, parents.csv is created, which looks like this:
Unlocode,Parent
DEBHQ,DEHAM
With this, you can easily find out these are related.
It's impossible to find out that both "Vienna" and "Wien" are in fact the same city with UN/LOCODE ATVIE. That is, if you use the official dataset.
Not so much with aliases-improved.csv, which looks like this:
Unlocode,Alias
ATVIE,Wien
ATVIE,Vienna
This is much more usable than the aliases in the original. Not only because of the improved user-friendliness, but mostly because of its sheer size. The official dataset has less than 100 aliases, this one has over 670,000.
The United Nations Code for Trade and Transport Locations is a code list maintained by UNECE (a United Nations agency) to facilitate trade. The list comes from the UNECE page, released at least once a year. However, this dataset is based on datasets/un-locode, which is already much better than the original (different CSVs for different purposes instead of cramming it all in 1).
CargoProbe offers the functionalities of this dataset as an API. This also has additional nifty things: like when you only have the city name "Rotterdam", it's clever enough to return the data of NLRTM instead of USRAJ.
Their main product is essentially track and trace for shipping containers, but they do a lot more. Check them out if that interests you.
All unlocode data is licensed under the ODC Public Domain Dedication and Licence (PDDL).
ODbL 1.0. http://osm.org/copyright
CC-0 (No rights reserved)