Most common address for Finnish buildings?

If my SQL is correct, the most common address name for a building in Finland is Rantatie (Shore road), if null values are excluded that is. According to my query*, there are a total of 5634 buildings residing along a road called Rantatie. The second most common address for a building is Kirkkotie (Church road), followed by Vanhatie (Old road), Koulutie (School road) and Keskustie (Central road).

rantatie.png

Rantatie buildings clustered.

Not the most interesting map, likely the buildings are where people are. Maybe calculating the proportion of Rantatie buildings per municipality or some such would be more interesting.

*Query below

SELECT osoite_fi, count(*)
FROM osoitteet.osoitteet_ei_null
GROUP BY osoite_fi
ORDER BY count(*) DESC
LIMIT 5
Mainokset

Mattolaiturit pt. II: does the carpet match the drapes?

Quite recently, the accessibility research group at the University of Helsinki published an updated version of their modified digiroad data. I had previously tried to use the data to calculate service/catchment areas, but ran into technical difficulties caused by my lack of SQL skills. But! I decided to give it a go once again.

I won’t go through the steps since Anita Graser has already done that. Some modifications to the SQL statements were necessary, and apparently some of the SQL functions have deprecated (?), too. So, below you can see the 10 minute service areas to the carpet washing places (places where you can wash your carpets, not sure what they’re called in English)  located in the Helsinki metropolitan area. The darker the red, the more time it takes to get to a carpet washing place. Areas outside the reds have dirty carpets, I assume.

image

I used the daily average travel time as the cost attribute. The end result is an interpolation of the vertices including the cost value. I am not really content with the use of IDW (had some issues with the other interpolation tools) but it will have to do for now. I cut the raster by using an alpha shape I produced earlier.

All in all, a nice exercise, although a bit burdensome to do. It would be nice to have a tool where you could just input the service locations as a point file and the tool would take care of the rest, provided that a working network data set is available. Still, it is amazing that the guys at the department of geology and geography produce these great data sets, not to mention the rest of you who created pgrouting tools and whatnot!