Title-1.gif (9040 bytes)

Ball.gif (922 bytes)Find Minimum Spanning Tree(for connected undirected graph)
            Tree : connected graph without cycle

                            iff.gif (149 bytes)forall.gif (91 bytes)u,v ,existonly.gif (146 bytes)  a path from u to v.

            Spanning Tree : edges connect all vertice of G

 Map1.gif (1146 bytes)

Ball.gif (922 bytes)Lemma

Map2.gif (2884 bytes)


           V1union.gif (115 bytes) V2union.gif (115 bytes) V3union.gif (115 bytes) V4 = V
           e = (V,W) minimum edge,v
belong.gif (100 bytes) Vi ,w~belong.gif (122 bytes) Vi
              
imply.gif (124 bytes) exist.gif (79 bytes)   MST include e
           proof : (T - {e'})
union.gif (115 bytes) e
                         is spanning tree of Lower cost.

Map3.gif (2395 bytes)

Ball.gif (922 bytes)Kruskal's algorithm
    forall.gif (91 bytes) vbelong.gif (100 bytes)V , {v} a component;                                          W(e1) lessequ.gif (102 bytes) W(e2)lessequ.gif (102 bytes) .......lessequ.gif (102 bytes)W(eN)
    repeat                                                                             for i = 1 to N do
            add minial edge across                
iff.gif (149 bytes)                       if add ei Not form cycle
                           2  components;                                                add ei
      Until  1  component remains

Ball.gif (922 bytes)Prim's algorithm
    V1 = visited

Map4.gif (1570 bytes)

    V2 = unvisited

   e = (v , w)
      = minimal edge  visited vertex 
assignr.gif (109 bytes)   fringe vertex
                                  

Map5.gif (2601 bytes)

Ball.gif (922 bytes)Priority - First - Search   implementation : Priority = Shortest length to tr