Exercises on Ch. 3

Minimum Spanning Tree Problem

  1. The MST Problem, whose name is the title of this chapter, pertains not to the basic "graphs" of previous chapters, whose edges are dichotomous (either present or absent, represented numerically by 0 and 1 respectively), but to a more general type of "networks" whose edges may have numerical values other than 0 and 1.

    Consider the network in Figure 3.1, page 56.  Construct a different network, which has the same nodes a through g, but all edge values changed from v to 13-v.  That is, edge ab changes from 12 to 1; edge af changes from 3 to 10; edge ag changes from 5 to 8; etc. Draw a diagram for this new network, labeling its edges with the new values.  We wish to find a MST for this new network.  

  2.  Repeat the MST construction of the previous exercise, except this time follow Boruvka's algorithm, first adding the smallest edge, then adding independent sets with increasing weights, then repeatedly connecting components using smallest edges.  

  3.  The theorems stated by Hage and Harary assert that the two algorithms reach the same end result, which is the MST for the network.  But the intervening steps were different.  Compare the two algorithms, in terms of what the MST-under-construction looked like (a) in exercise 2, at the stage when only Boruvka's independent sets had been added; and (b) in exercise 1, when the same number of edges as in part (a) had been added.