GeoNames Forum

I've found 193 cities in the cities1000.txt file that have 0 population. On a Unix/Linux system, you can reproduce the results with the following command:
cut -f1,3,15 cities1000.txt | egrep '[^0-9]0$'

There are an additional 21 cities that have populations between 0 and 1000.

I assume that this is a bug in the code that creates the cities1000.txt file?

it is not a bug, just missing data. Feel free to help add it: http://www.geonames.org/manual.html

Marc

"it is not a bug, just missing data"
I could understand this argument if I were talking about the _whole_ geonames database. However, I am referring to the database dump for those cities that are supposed to have a population of 1000 or greater. How do you know those cities have a population of 1000+ if the population field is 0?

They are capitals of states or provinces and it is to be assumed they have larger population. And even if they should have slightly less than 1000, they still belong to the list as being of some relevance.

Marc

Ok, then I think I would change the documentation that describes that file. Currently it says:
"all cities with a population > 1000"

Operating on the principle of least surprise, I assume that means that all cities I encounter in the file should have a population > 1000.

Thanks for the response,

Mike

does this mean for the allcountries zip that has some 2 million + cities that all extra cities added differently then 1000.zip have less then 1000 as a population.

I guess to make it clearer .. does this mean there are over 2 million cities with less then 1000 as a population ??

No, this thread speaks about places that have 'NO population data' in the geonames database. If the population information is missing then it does not necessarily mean that the population is lower than 1000.

Best

Marc

PS: I have deleted your other thread, it is not required to post two identical questions in different threads.