I have the feeling this will be more for posterity than anything else, but I ran a script which compares the coordinates of the unlocodes with easy_allCountries.csv
All in all, there were: 3481 who are within 250 km. 2607 who weren't.
If I just check the first 10 differences:
Instances where your CSV is more accurate:
ASPPG
ARPRQ
Instances where the unlocode location is more accurate:
ARSAA
ARSAN
ARSAO
ARSAR
ARSAT
ARSZQ
ATLOI
AUABC
That ain't a great score. So I guess just using the unlocode locations yields more accurate coordinates (even though that's inaccurate too: finding the 2 inaccurate locations in unlocode was way too easy).
Surprisingly,
perfect_allCountries.csv seems to be much more accurate (Close: 13453. Far: 384)
good_allCountries.csv is also good (Close: 50403. Far: 4878)
I have the feeling this will be more for posterity than anything else, but I ran a script which compares the coordinates of the unlocodes with easy_allCountries.csv
The output of the first 100 lines:
All in all, there were: 3481 who are within 250 km. 2607 who weren't.
If I just check the first 10 differences: Instances where your CSV is more accurate:
Instances where the unlocode location is more accurate:
That ain't a great score. So I guess just using the unlocode locations yields more accurate coordinates (even though that's inaccurate too: finding the 2 inaccurate locations in unlocode was way too easy).
Surprisingly, perfect_allCountries.csv seems to be much more accurate (Close: 13453. Far: 384) good_allCountries.csv is also good (Close: 50403. Far: 4878)