Basics

Definition

A hypergraph is a set of nodes and a collection of hyperedges, each connecting any number of nodes.

You will learn

Build a small hypergraph and use the core HGX API for inspection and basic measures.

Overview

  • Build a first hypergraph and inspect nodes, edges, and metadata.

  • Compute a few core measures to get comfortable with the API.

Setup

[1]:

import matplotlib as mpl

mpl.rcParams.update({
    "figure.figsize": (6, 4),
    "figure.dpi": 120,
    "savefig.dpi": 150,
})

[2]:

import sys
sys.path.append("..")

from hypergraphx.generation.scale_free import scale_free_hypergraph
from hypergraphx.linalg import *
from hypergraphx.representations.projections import bipartite_projection, clique_projection
from hypergraphx.generation.random import *
from hypergraphx.readwrite.save import save_hypergraph
from hypergraphx.readwrite.load import load_hypergraph
from hypergraphx.viz.draw_hypergraph import draw_hypergraph

[3]:

H = Hypergraph([(1, 3), (1, 4), (1, 2), (5, 6, 7, 8), (1, 2, 3)])
print(H)

Hypergraph with 8 nodes and 5 edges.
Distribution of hyperedge sizes: {2: 3, 4: 1, 3: 1}

[4]:

H = Hypergraph([(1, 3), (1, 4), (1, 2), (5, 6, 7, 8), (1, 2, 3)], weighted=True, weights=[1, 2, 3, 4, 5])

[5]:

H.get_edges()

[5]:

[(1, 3), (1, 4), (1, 2), (5, 6, 7, 8), (1, 2, 3)]

[6]:

H.get_weights()

[6]:

[1, 2, 3, 4, 5]

[7]:

for edge in H:
    print(edge)

((1, 3), 0)
((1, 4), 1)
((1, 2), 2)
((5, 6, 7, 8), 3)
((1, 2, 3), 4)

[8]:

H.get_weight((1, 2, 3))

[8]:

5

[9]:

H.get_weight((2, 3, 1))

[9]:

5

[10]:

H.is_connected()

[10]:

False

[11]:

draw_hypergraph(H)

[11]:

<Axes: >
../_images/tutorials_basics_15_1.png 
[12]:

H.add_edge((1,5), weight=10)

[13]:

draw_hypergraph(H)

[13]:

<Axes: >
../_images/tutorials_basics_17_1.png 
[14]:

H.is_connected()

[14]:

True

[15]:

H.is_uniform()

[15]:

False

[16]:

print(H.check_edge((5, 6, 7, 8)))
H.remove_edge((5, 6, 7, 8))

True

[17]:

print(H.get_edges())
print(H.get_weights())
H.remove_edges([(1,2,3), (1,3)])

[(1, 3), (1, 4), (1, 2), (1, 2, 3), (1, 5)]
[1, 2, 3, 5, 10]

[18]:

print(H.get_edges())
print(H.is_uniform())

[(1, 4), (1, 2), (1, 5)]
True

[19]:

H.get_edges()

[19]:

[(1, 4), (1, 2), (1, 5)]

[20]:

H.add_attr_to_node_metadata(1, 'role', 'student')

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[20], line 1
----> 1 H.add_attr_to_node_metadata(1, 'role', 'student')

AttributeError: 'Hypergraph' object has no attribute 'add_attr_to_node_metadata'

[ ]:

H.get_nodes(metadata=True)

{1: {'role': 'student'}, 3: {}, 4: {}, 2: {}, 5: {}, 6: {}, 7: {}, 8: {}}

[ ]:

H.remove_node(1)

[ ]:

H.get_nodes(metadata=True)

{3: {}, 4: {}, 2: {}, 5: {}, 6: {}, 7: {}, 8: {}}