[OSM-talk-be] more stats: data density in the Belgian regions

Tue Nov 10 07:17:06 UTC 2015

Marc,

This is not impossible to achieve. For example, I almost finished an
analysis once to detect which people are/have been mapping biking related
stuff in the past. In my grand scheme of things, I also intend to do
classification of things by group, like you suggested. I'll release some of
these stats about Belgium in the coming days/weeks. Biggest issue: my
processing doesn't do relations, so I do miss some interesting things.

As to the thing you want to do with similar mappers: does it have to be a
history file? I'm not sure, and if it isn't, it's probably easier to do it
on a snapshot. That would make regular updating also easier.
That said, I can release some of my intermediate files as tables or csv,
someone might be able to make some kind of website out of that. But that
would again be with a local scope, as I don't have the capacity to process
global files at once. (I cut up the world in little pieces to run analysis,
but that means you need to finish a processing script before rolling it
out, and I'm not good at finishing things)
Some of the questions you asked have little to do with local context, so it
might be more interesting to see how a thing like taginfo works and build
upon that.

My personal interest is more about "map completeness" for road networks,
landuse, amenities, etc; with a global scope. In the second place mapper
inequality and remote mapping. For things like that, I don't see another
approach than taking world history and cutting it in pieces...

2015-11-10 7:43 GMT+01:00 Marc Gemis <marc.gemis at gmail.com>:

> Thanks Joost,
>
> It's always good to see such graphs. I'm interested to learn
>
> * What are we mapping now ?
> * What did we map in the past ?
> * Who is mapping what  ?
> * Who is mapping similar things as me ? (Or am I the only one mapping
> feature X ? / can I ask someone also for help with feature X ?)
> * Where are we mapping feature X (on town/village level) ?
>
>
> With "what" I mean e.g. categories of features: streets (highway tags),
> landuse/landcover, boundaries, buildings, shops, etc.
> However, for some purposes categories might be broken. mapping
> amenity=library or amenity=bench are quite different. The same holds for
> tourism=hotel and tourism=information,information=board. The library and
> hotel are "important" features, bench or an information board not so. I
> know "important" is relative, but I hope you understand what I mean.
>
> Of course this is harder when we start thinking about "attributes": turn
> lanes, destinations, house numbers, or additional attributes for
> amenity=bicycle_parking (such as covered or bicycle_parking=...), etc.
>
> For the last two questions (mapping feature X) the ultimate goal would be
> to answer questions such as "Who else is mapping bicycle_parkings ?" With
> which attributes. ? Where is feature X not mapped at all ?
>
> Regards
>
> m
>
>
> On Mon, Nov 9, 2015 at 8:59 AM, joost schouppe <joost.schouppe at gmail.com>
> wrote:
>
>> Hi all,
>>
>> So, Nicolas' remark after my last post got me interested again in some
>> stats about OSM data in Belgium. Is there any difference between Brussels,
>> Wallonia and Flanders?
>> Of course, I made a little mistake and my laptop will have to redo some
>> calculations this night, but here's something that worked.
>>
>> How did data density evolve in the three parts of the country over the
>> last years?
>>
>> This is a simple count of nodes that were in existence on january first
>> of any year. So that includes independent nodes, but also nodes used (and
>> used again and again) for building lines and relations.
>> To make the regions comparable, I standardized by population. The idea is
>> that most stuff that we map is a function of humans, no so much of area.
>> (Yes, I know, same population in a larger area would probably imply more
>> things to map)
>>
>> So this graph shows the evolution of nodes per 1000 of population.
>> Flanders was clearly highest since 2010. Wallonia started of much quicker
>> than Brussels, but can't keep up with Flanders. In Brussels we have a very
>> obvious jump in 2014. That's probably the buildings/addresses import.
>>
>> http://i.imgur.com/RPK38DM.jpg
>>
>>
>> Next thing I want to do is see how many different mappers have built the
>> map.
>>
>> I just stumbled upon the very first nodes and lines in Flanders, and the
>> user is still active. The story of these nodes is in his diary, very bottom
>> of the page:
>> http://wiki.openstreetmap.org/wiki/User:LA2/Diary_for_Q3_2005
>>
>>
>> --
>> Joost @
>> Openstreetmap <http://www.openstreetmap.org/user/joost%20schouppe/> |
>> Twitter <https://twitter.com/joostjakob> | LinkedIn
>> <https://www.linkedin.com/pub/joost-schouppe/48/939/603> | Meetup
>> <http://www.meetup.com/OpenStreetMap-Belgium/members/97979802/> | Reddit
>> <https://www.reddit.com/u/joostjakob> | Wordpress
>> <https://joostschouppe.wordpress.com/>
>>
>> _______________________________________________
>> Talk-be mailing list
>> Talk-be at openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-be
>>
>>
>
> _______________________________________________
> Talk-be mailing list
> Talk-be at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-be
>
>

-- 
Joost @
Openstreetmap <http://www.openstreetmap.org/user/joost%20schouppe/> |
Twitter <https://twitter.com/joostjakob> | LinkedIn
<https://www.linkedin.com/pub/joost-schouppe/48/939/603> | Meetup
<http://www.meetup.com/OpenStreetMap-Belgium/members/97979802/> | Reddit
<https://www.reddit.com/u/joostjakob> | Wordpress
<https://joostschouppe.wordpress.com/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-be/attachments/20151110/21798e45/attachment.htm>