|
Is it possible to query for a tag without downloading OSM data files? I would like to be able extract some statistics on how often the wheelchair tag is used, by country, without having to download osm data for the entire world. |
|
taginfo can provide a good overview on the usage of tags. For example you can see on which elements (nodes, ways, relations) the wheelchair tag is used, typical values and their occurrence, combinations with other tags and there is also a simple world map showing the tag distribution. 1
But that misses the "by country" which is explicitly mentioned in the question.
(16 Aug '13, 12:42)
aseerel4c26 ♦
1
Taginfo looks helpful. I will check out the downloadable data. I am trying out OSM and open source GIS tools for the first time, so I will need some time to see if I can get the answer. If the data has coordinates, then I think I should be able to determine the country using the other tools.
(16 Aug '13, 16:46)
Cstudent
2
There is at least one country-specific instance of taginfo: http://taginfo.openstreetmap.org.uk/ Not sure how many others there are though.
(17 Aug '13, 10:01)
SomeoneElse ♦
4
I previously didn't know there where some TagInfo sites available per country. A big thanks to the individual maintainers. A list of known instances is kept here:
(18 Aug '13, 11:40)
robbieonsea
|
|
I just found overpass turbo. That looks very helpful. |
|
The opposite of the taginfo tool which scai mentioned: You can download extracts (e.g. by country) of our data - so you do not need to download the "whole planet". On that data extract you could do your analyses (who often a specific tag is used, …). I am having trouble finding what I need there. Perhaps you can direct me to an appropriate extract. I have given up trying to get data for the whole world. I would be satisfied with just the US. I would like get an extract of all nodes in the US containing the following information: name, wheelchair, latitude, longitude. I will need to normalize the data somehow. I was thinking that for each area (state or county) I would calculate the percent of named nodes that have the wheelchair tag. If anybody has a better idea, I would be happy to hear it.
(17 Aug '13, 03:24)
Cstudent
Well, I have never worked with the raw data... and so I have not really a clue if that is the best way to go for you. However, via the wiki page I had linked to, you get to e.g. http://download.geofabrik.de/north-america.html which offers you the US in four parts (not as a whole apparently; scroll down).
(17 Aug '13, 03:49)
aseerel4c26 ♦
2
If you download an extract of the data you've got a couple of options of how to process it. One is to download as a .osm.gz file and uncompress it to a raw xml .osm file. Then you can use standard command-line utilities like grep etc. to search for things in that data (have a look here if you're on windows for how to get grep etc. for that platform). The downside to that approach is that .osm files for North America will be very big indeed (many 10s of Gb). Another is to download as a .pbf file and process that. The downside here is that you'll need something that can process the file format - either write something yourself or use one of the tools on this list. Of those, Osmosis may be the place to start - it has lots of functionality (including geographical selection) and you may be able to pair it with osmconvert and osmfilter to count things in a certain area.
(17 Aug '13, 10:18)
SomeoneElse ♦
Thanks for the information. I should have said that I am working from home - slow internet and not very powerful computer. I think that this will not be a good path under those circumstances. I have a bittorrent download in progress. It ran all night and it still has not finished. I suspect that I will have trouble processing the file when that has finished. I have a Mac and use open source GIS tools on an Ubuntu VM with 1 Gb RAM.
(17 Aug '13, 18:01)
Cstudent
|