Pytroids

---

A small Python package to perform computations on graphic matroids. It is compatible with NumPy (NumPy adjacency matrices as inputs) and NetworkX (NetworkX graphs as inputs).

• Free software: MIT license
• Documentation: TODO

Examples

import numpy as np
import networkx as nx
from pytroids.graphic_matroids import *
from pytroids.dual_matroids import *

n = 10
p = 0.2
G = nx.gnp_random_graph(n, p, seed=None, directed=False)
A = nx.adjacency_matrix(G)

#Input: NumPy adjacency matrix:
M = GraphicMatroid(A)

print("Edges:", M.edges)
print("Rank:", M.rank(M.edges))
print("Bases:", M.bases())
print("Circuits:", M.circuits())

#Input: NetworkX graph:
M = GraphicMatroidNX(G)

print("Edges:", M.edges)
print("Rank:", M.rank())
print("Bases:", M.bases())
print("Circuits:", M.circuits())

#Dual matroid:
M_star = DualMatroid(M)

print("Primal bases:", M.bases())
print("Dual bases:", M_star.bases())
print("Dual rank:", M_star.rank())

Warnings

This package does not implement any optimized algorithm for computing graphical matroids. Therefore, it is only efficient for minimal graphs with a few nodes.

Dependencies

• NumPy
• NetworkX