English

Area, Perimeter, Height, and Width of Rectangle Visibility Graphs

Combinatorics 2022-08-29 v2

Abstract

A rectangle visibility graph (RVG) is represented by assigning to each vertex a rectangle in the plane with horizontal and vertical sides in such a way that edges in the graph correspond to unobstructed horizontal and vertical lines of sight between their corresponding rectangles. To discretize, we consider only rectangles whose corners have integer coordinates. For any given RVG, we seek a representation with smallest bounding box as measured by its area, perimeter, width, or height (height is assumed not to exceed width). We derive a number of results regarding these parameters. Using these results, we show that these four measures are distinct, in the sense that there exist graphs G1G_1 and G2G_2 with area(G1)<area(G2)area(G_1)<area(G_2) but perimeter(G2)<perimeter(G1)perimeter(G_2)<perimeter(G_1), and analogously for all other pairs of these parameters. We further show that there exists a graph G3G_3 with representations S1S_1 and S2S_2 such that area(G3)=area(S1)<area(S2)area(G_3)=area(S_1)<area(S_2) but perimeter(G3)=perimeter(S2)<perimeter(S1)perimeter(G_3)=perimeter(S_2)<perimeter(S_1). In other words, G3G_3 requires distinct representations to minimize area and perimeter. Similarly, such graphs exist to demonstrate the independence of all other pairs of these parameters. Among graphs with n6n \leq 6 vertices, the empty graph EnE_n requires largest area. But for graphs with n=7n=7 and n=8n=8 vertices, we show that the complete graphs K7K_7 and K8K_8 require larger area than E7E_7 and E8E_8, respectively. Using this, we show that for all n8n \geq 8, the empty graph EnE_n does not have largest height, width, area, or perimeter among all RVGs on nn vertices.

Keywords

Cite

@article{arxiv.2208.09961,
  title  = {Area, Perimeter, Height, and Width of Rectangle Visibility Graphs},
  author = {John S. Caughman and Charles L. Dunn and Joshua D. Laison and Nancy Ann Neudauer and Colin L. Starr},
  journal= {arXiv preprint arXiv:2208.09961},
  year   = {2022}
}
R2 v1 2026-06-25T01:51:16.475Z