Hi All,
I have been puzzling over the TSP matrix option and its use for non-Euclidean distances in large part because the routing problem I normally solve is not a rational city layout, but an irrational trail layout which regularly violates the expectation of Euclidean distance (places like this: http://maps.clemetparks.com/url/fjc or even better this: http://maps.clemetparks.com/url/cjc .
So, am I correct to assume that the matrix itself is like the distance matrices in the back of an old atlas, or am I way off here?
Chicago Denver New York
Chicago 0 1004 790
Denver 1004 0 1778
New York 790 1778 0
Thanks,
Best,
Steve
On 6/15/2013 11:17 AM, Stephen Mather wrote:
Hi All,
I have been puzzling over the TSP matrix option and its use for
non-Euclidean distances in large part because the routing problem I
normally solve is not a rational city layout, but an irrational trail
layout which regularly violates the expectation of Euclidean distance
(places like this: http://maps.clemetparks.com/url/fjc or even better
this: http://maps.clemetparks.com/url/cjc .
So, am I correct to assume that the matrix itself is like the distance
matrices in the back of an old atlas, or am I way off here?
Chicago Denver New York
Chicago 0 1004 790
Denver 1004 0 1778
New York 790 1778 0
Yes this is exactly what it is.
-Steve