Drawing Large Graphs with a Potential-Field-Based Multilevel Algorithm

Large Quasi-Tree Drawing: A Neighborhood Based Approach

“…In term of computation time, the algorithm FM 3 [16] is clearly the most efficient over all these algorithms. It is 10 times faster on Net05 and more than 20 times faster on pgraph than our approach.…”

“…These results are compared to results obtained by other algorithms. We decided to compare our results to FM 3 of [16], LGL of [2] and to SPF of [3]. The choice of these three algorithms is dictated by three different criteria: first of all, the computation time (FM 3 of [16]), then the esthetic criteria (LGL of [2]), and finally, the "tree-like" property emphasizing (SPF of [3]).…”

“…Among them, the most famous is based on improvements of force directed methods ( [15,16,2]). Another approach consists in representing the graph as a matrix and using linear algebra techniques ( [18,27]).…”

“…Hachul and Jünger describe in [16] another vertex filtration in which the graph is considered as a set of solar systems composed of suns, planets and moons. Then, each solar system is collapsed into a single vertex and that process is repeated to produce a serie of graphs G = G 0 , G 1 , ..., G k , where G k is considered as "small" enough.…”

Drawing Clustered Graphs as Topographic Maps

“…Force-directed layout is widely used. After more than two decades of development, there exist algorithms and according software tools that, based on the original idea, can deal with large graphs, see, e.g., Hachul et al [16].…”