I wonder why N_G(F) costs O(|N_G(F)|log|F|) work and O(log^2|F|) span.
I think, since merge A B has O(|A|+||B|) work
and O(log(|A|+|B|)) span, N_G(F) costs differently.
(I'm not sure, bit maybe O(|N_G(F)|×|F|) work and O(log(|N_G(F)|×log|F|) span.
Plus, do we need to study about the work efficient merge algorithm that has O(logN) span?
The above figure is from the previous answer.
I wonder why N_G(F) costs O(|N_G(F)|log|F|) work and O(log^2|F|) span.
I think, since merge A B has O(|A|+||B|) work and O(log(|A|+|B|)) span, N_G(F) costs differently. (I'm not sure, bit maybe O(|N_G(F)|×|F|) work and O(log(|N_G(F)|×log|F|) span.
Plus, do we need to study about the work efficient merge algorithm that has O(logN) span?