On 19/01/16 15:25, Robert Whittaker (OSM lists) wrote: > On 19 January 2016 at 13:17, Bogus Zaba <bog...@bogzab.plus.com> wrote: >> At the end of the day however, if Welsh schools are tagged with >> ref:estyn=* but not with ref:edubase=*, am I right in thinking that they >> will not be picked up as "progress" in the quarterly project? > All UK ways and relations tagged with either amenity=school or > amenity=college will get picked up by my tool at > http://robert.mathmos.net/osm/schools/ . I then attempt to match these > OSM objects with the schools on the official lists. Any ref:edubase, > ref:seedcode and ref:deniirn values are used on a first pass, but > after that there is a second pass that matches each unmatched school > on the official list to the nearest OSM object within 1km that hasn't > yet been matched. In rural areas (schools generally more than 1km > apart) this works very well. In urban areas the tool can often match > things the wrong way. But a match will still correspond to a school in > OSM being counted, even if it's an incorrect match. > > So apart from some of the matches in the tool being wrong, almost > everything should still be counted. (The exception is if there's a > missing postcode in the official data, or the actual location of the > school is more than 1km from the postcode centroid. Then the match > won't happen.) > > Hope that helps, > > Robert. > OK, I get that now. I'll crack on with what I was doing last week.
Main problems I am seeing is with schools apparently changing names (is it just a name change?) or adding new school on the same site as houses another one. Some will require ground surveys which I will not be able to do as I draw circles of greater and greater circumference from home base. Bogus -- Dr Bogumil N Zaba _______________________________________________ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
