what is the running time of dfs if we represent its input graph by an adjacency matrix and modify the algorithm to handle this form of input? hint: the algorithm we looked at and analyzed in class was for an adjacency list. how would this change for an adjacency matrix?