It doesn’t make sense to make threshold bigger than the maximal weight, as it’s going to include all the edges anyways.
The complexity, m^2 \log c, also seems wrong to me (picking c=1 doesn’t seem like the best idea), but it isn’t immediately obvious what’s the right estimate here.
It doesn’t make sense to make threshold bigger than the maximal weight, as it’s going to include all the edges anyways.
The complexity, m^2 \log c, also seems wrong to me (picking c=1 doesn’t seem like the best idea), but it isn’t immediately obvious what’s the right estimate here.