Files
math-research/plane_depth_sequencing.py
T

66 lines
2.7 KiB
Python

"""Plane depth sequencing on maximal planar graphs."""
import random
from typing import Any, TypedDict
from sage.all import Graph, graphs # type: ignore[attr-defined] # pylint: disable=no-name-in-module
class DeeplyEmbeddedGraph(TypedDict):
graph: Graph
outer_cycle: list[Any]
plane_depth_labelling: dict[Any, int]
deep_embedding: Graph
def get_plane_depth_labelling(g: Graph, outer_cycle: list[Any]) -> dict[Any, int]:
"""Return the plane depth of each vertex relative to the given outer cycle."""
# equivalent to the commented out naive implementation:
# return {v: min(g.distance(v, u) for u in outer_cycle) for v in g.vertices()}
source = max(g.vertices()) + 1
g_prime = g.copy()
g_prime.add_vertex(source)
g_prime.add_edges([(source, v) for v in outer_cycle])
distances = g_prime.breadth_first_search(source, report_distance=True)
return {v: d - 1 for v, d in distances if v != source}
def deep_embedding(g: Graph, outer_cycle: list[Any], plane_depth_labelling: dict[Any, int] | None = None) -> Graph:
"""
Return the deep embedding of g relative to outer_cycle.
For every interior triangular face whose three vertices all have the same
plane depth, a new vertex is added adjacent to each vertex of that face.
"""
if plane_depth_labelling is None:
plane_depth_labelling = get_plane_depth_labelling(g, outer_cycle)
outer_vertices = set(outer_cycle)
embedding = g.get_embedding()
if embedding is None:
g.is_planar(set_embedding=True)
embedding = g.get_embedding()
g_prime = g.copy()
new_vertex = max(g.vertices()) + 1
for face in g.faces(embedding):
face_vertices = [u for u, v in face]
if set(face_vertices) == outer_vertices:
continue
if plane_depth_labelling[face_vertices[0]] == plane_depth_labelling[face_vertices[1]] == plane_depth_labelling[face_vertices[2]]:
g_prime.add_vertex(new_vertex)
g_prime.add_edges([(new_vertex, v) for v in face_vertices])
new_vertex += 1
return g_prime
def generate_example(n: int) -> DeeplyEmbeddedGraph:
"""Generate a random maximal planar graph of size n and return the triangulation, outer cycle, and deep embedding."""
g = graphs.RandomTriangulation(n)
g.is_planar(set_embedding=True)
embedding = g.get_embedding()
faces = g.faces(embedding)
outer_cycle = [u for u, v in random.choice(faces)]
plane_depth_labelling = get_plane_depth_labelling(g, outer_cycle)
return DeeplyEmbeddedGraph(graph=g, outer_cycle=outer_cycle, plane_depth_labelling=plane_depth_labelling, deep_embedding=deep_embedding(g, outer_cycle, plane_depth_labelling))