GeoNames Forum

For Spain we are using the following administrative divisions :

ADM1 = comunidades autónomas (17)
ADM2 = provincias (50+2)
ADM3 = comarcas / municipios (*)
ADM4 = municipios

* : In Catalunya we have four administrative divisions with comarcas as ADM3 and Municipios as ADM4. In all other comunidades we have only three levels.

Shouldn't all "municipios" in Spain (feature code ADM3 or, for municipios in Catalunya, ADM4) exist as populated places too (feature code PPL)?

More specifically, I've noticed that a municipio called Zigoitia ( http://en.wikipedia.org/wiki/Zigoitia ) only appears as an ADM3 administrative division
http://www.geonames.org/search.html?q=zigoitia&country=ES
Should I add it as a populated place (PPL) too? (like it happens with other municipios, like Labastida, which have 2 entries, one as ADM3 and the other as PPL).

Thanks.

xavi

Xavi

If there is a populated place with the name of "Zigoitia" then there should be an entry for it. On googlemaps I cannot see anything there that looks like a populated place and if it is the name of the municipality put their is no populated place with this name then there should be no entry.

Marc

Thanks for the response Marc.

I think my confusion comes from the term "populated place". I thought of it as the smallest area with a mayor and a city council. According to that definition, Zigoitia would be a populated place (see the Spanish Wikipedia for more info about Zigoitia, http://es.wikipedia.org/wiki/Cigoitia ). But from your response, and re-reading the definition of the PPL feature code ("a city, town, village, or other agglomeration of buildings where people live and work", http://www.geonames.org/export/codes.html ), it might be that the essential thing about populated places is that they are "agglomerations"? I mean, if the smallest area with a mayor and a city council doesn't concentrate all the population in the same place (i.e. it isn't in the same agglomeration), but it's dispersed in several groups of houses across the area (like it seems it's the case for Zigoitia), then it's not a populated place?

Sorry if I'm being a pain, but I'm still confused about the meaning of "populated places" in Geonames.

xavi

'PPL' stands for larger groups of buildings where people are permanently living. The administrative and political categorization is modelled with the featureclass 'A'. If some dispersed groups of houses share some administrative and political infrastructure and build a municipality then I wouldn't call them populated place. The entry of featureclass 'A' already expresses the idea of the municipality.
Each group of houses that has a name would get an entry of featureclass 'P', depending on the size a PPL, PPLL or PPLX.

Best

Marc

marc wrote:

For Spain we are using the following administrative divisions :

ADM1 = comunidades autónomas (17)
ADM2 = provincias (50+2)
ADM3 = comarcas / municipios (*)
ADM4 = municipios

* : In Catalunya we have four administrative divisions with comarcas as ADM3 and Municipios as ADM4. In all other comunidades we have only three levels.

Hi marc !

First of all, I'd like to thank you (and the community) for the incredible work done on geonames.

About this municipios, wouldn't it make sense that all municipior in Spain have an ADM4 entry, with an empty ADM3 in case they are not in Catalunya ?
Maybe 'having ADM4 without ADM3' breaks your data model, but I find it quite difficult to work on this dataset where the same data (i.e. administrative code) is spread over two columns.
(I have a temp table with both columns merged, but it would still be a lot easier for updates & all if the data was in the same column)

Regards

I agree that this is not an optimal solution. I think it was wrong to include the comarcas at all. It would be easier without them. Adding a 'null' level between admin levels is also creating problems.
I guess we should move the adm4 to adm3 and move the comarcas to ADMD. we would loose the hierarchy information of the comarcas with this move.

Marc

Example Ceuta: ES;00;51

First level division: '00' stands for 'not applicable' (http://forum.geonames.org/gforum/posts/list/448.page#7005)

Sugestion: Put the first level division as ADM1, and '00' for second if not aplicable. Or null.

Ceuta: ES;51;
As you do with AR;07;

This way you can add adm4,5,6,7...

Or "first order administrative division" is not the same that "first level division"

Regards

The problem here is that Ceuta does not have a first division code, so we can put the adm2 code into the adm1 column. When we change the first code from the fips codes to the iso codes we could use the iso code for Ceuta, but I don't want to change just one country, we should coordinate it and move as many countries as possible. It would solve the adm1 issue in this case, but there will be other situations where iso does not have assigned a code.

Marc

And what about HASC code?
http://www.statoids.com/ues.html
http://www.statoids.com/uus.html

The hasc codes are only used on the statoids website and are not an official and commonly used coding system like fips or iso. In situations where do not have a code at all we use the geonameId.

Marc

For Spain, INE base uses 18 for Ceuta and 19 for Melilla.
Latest update was on February 3:
http://www.ine.es/jaxi/menu.do?type=pcaxis&path=%2Ft20%2Fe245%2Fcodmun%2F&file=inebase&L=0

Hello,

Spain has 19 first class administrative divisions, not 17.

I suggest to use the ISO codes instead of the FIPS codes, because there are not FIPS codes for Ceuta and Melilla.

http://en.wikipedia.org/wiki/ISO_3166-2:ES

There is a problem in the file admin1Codes.txt

Melilla is not in the list but there is this obviously wrong line:

Code:

ES.28	(SP28)

Thank you very much
Alfredo

alfredono wrote:

Spain has 19 first class administrative divisions, not 17.[/code]

Since 1983 there were 17 autonomous regions created, but you are talking about the 2 autonomous cities which, yes then it is 19.