-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi,

after struggling a few days with the concept of tower/pillar nodes, or
rather how they are defined during reading the OSM file, I think I
have finally understood how it works…

As I think I’m not the only one who did not get this at the first
grasp, I want to improve the code to better express what is really
done (that is – in my understanding – count how often each node was
seen and if it is encountered at least twice, mark it as a tower node).

For that, I want to create two new classes/interfaces: a NodeMap, that
encapsulates the source (OSM)->internal ID mapping, and a NodeTypeMap
that uses the NodeMap to get the node type. The latter can probably be
thrown away after preprocessing, I did not think that through till the
end.

What do you think of that approach? Is it a good idea? Or am I missing
a part of the bigger picture that makes this unfeasible?


Regards
Andreas

PS: This being my first mail to this list, I should probably shortly
introduce myself: I’m studying Electrical Engineering and Information
Technology at KIT (fka University of Karlsruhe). Currently I’m writing
my diploma thesis, using GraphHopper as a tool to generate driving cycle
s.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iF4EAREIAAYFAlVJ4GkACgkQRJlqDhrZfp2gsQD/XnJOHHKJ0a3n5LokkqFwGimB
BH/V++AQFN8F6HCSzFkBAIQWHIxVP2qgA65R4C4SWISX91iZMlDx+Q0EcEkI9dR+
=xbgs
-----END PGP SIGNATURE-----

_______________________________________________
GraphHopper mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/graphhopper

Reply via email to