Skip to main content
7 events
when toggle format what by license comment
Jan 16, 2021 at 11:04 comment added Massimo Cafaro What you mean exactly by "clean comments" ?
Jan 16, 2021 at 7:26 vote accept Okh. Pij
Jan 7, 2021 at 9:55 comment added Massimo Cafaro Your counter-example is wrong: in $k$ iterations the algorithm is only considering paths of length $k$. Try running BF and you will se that after 2 iterations only the paths $s$-->$t$ and $s$-->$u$ have been correctly computed (the distance for $s$-->$v$ will still be $\infty$).
Jan 6, 2021 at 17:14 comment added Massimo Cafaro The link you provide has two questions. The first one is exactly the one which I have answered, providing full details: besides the standard induction proof there is nothing that can be really usefully added. The second claim, the one related to the weights on a shortest paths is obviously false, as already shown in the answer to the question you linked to. Regarding the variants available in Jeff Erickson's book, the first is just a minor variation due to Moore (which is faster) and the other three are just the steps needed to simplify the original algorithm based on dynamic programming.
Jan 6, 2021 at 15:09 comment added Massimo Cafaro in your original question there is no reference to a modified algorithm. The answer refers to the Bellman Ford original algorithm. By the way, the example given in your comment can not be understood, since you are not providing the full input graph and the weights on each of the edges.
Dec 9, 2020 at 7:37 history edited Massimo Cafaro CC BY-SA 4.0
added 71 characters in body
Dec 8, 2020 at 11:33 history answered Massimo Cafaro CC BY-SA 4.0