From: Steffen on
Is there any other resource on the topic of finding the k shortest
paths in digraph than the one publicized by David Eppstein but using
the same algorithm? I read the article
"http://citeseer.ifi.unizh.ch/cache/papers/cs/12009/http:zSzzSzwww.ics.uci.eduzSz~eppsteinzSzpubszSzEpp-SJC-99.pdf/eppstein97finding.pdf"
but I think I need more "skills" to understand all of the basic
algorithm described there. Or, maybe someone can help me understanding
it: What is the acyclic graph D(G) and the graph P(G) about? (I would
prefer useful examples in this case)

Continue hoping to understand,
Steffen

From: Steffen on
While reading again the article mentioned above I wondered about the
sample path tree at page 8: Why doesn't have the root of the path tree
a child {9}? The edge 9 in G - T obviously has its tail at the path
from s to t in T, so I don't know why it is missing there.
And a second thing: It is possible to use breadth-first-search in that
path tree to get the k shortest pathes? Eppstein means not (page 7),
but if we see the path tree as m-heap and use the sum of the
delta(e)-values of a node (which is the set sidetracks(p)) as
heap-order, then this heap contains as root always the next shortest
path. Is that right?

Thanks for any help,
Steffen