Categories of "relation status" in xDima2

[Bernd Welter]

Hi folks,

for demonstration purposes I calculated a dima with some special coinstraints:

• 5 locations somewhere on the German island Sylt
• A custom bounding rectangle which "excludes" the point in the far east
• Violations allowed

I wanted to demonstrate that the dima is complete (5x5 relations) and splits the relations into 3 categories:

• relation routed successful, without violations
• relation routed successful, with violations
• relation not routed successful, therefore estimated by direct distance

What surprises me: the point in the far east contributes with:

• 3 times without violations
• 2 times violated
• 3 times estimated by direct distance

How can this be? I expected the location to be "unreachable" from a symmetric perspective.
Bernd

[Bernd Welter]

OK, that was a tricky one - probably doesn't occur in reality

• The isolated coordinate in the east (green pin) is linked to the closest segment (green star) that touches the rectangle's border.
• That particular road is a oneway street which
  • enables the router to leave the place
  • but prevents him from reaching the eastern point

Thanks for the analysis, Frank!

That being said I recommend that you also check this article "xServer2: Handling of optimization on islands such as Sylt, Helgoland, ..." carefully if you are interested in "spatial" challenges

Bernd