Graph of objects using graph-tool

Question

I'm trying to use the library graph-tool to create a new graph of objects. In particular I want some nodes of students and others of professors.

In the following the code I've written, I'm not able to add the vertices of students/professors.

The documentations https://graph-tool.skewed.de/static/doc/quickstart.html#property-maps explains that I have to use the class PropertyMap, it's quite weird for me...

Could someone explain to me how to use it? Thanks.

This is my code:

from graph_tool.all import *

class Student:
  def __init__(self, name):
    self.name = name

class Professor:
    def __init__(self, name):
         self.name = name

g = Graph()
s = Student("Luke")
p = Professor("Adam")
print(s.name)
print(p.name)

score 0 · Answer 1 · edited Oct 27 '19 at 21:44

Well first you need to add the vertices to the graph

# your code
# from graph_tool.all import *  # avoid importing all, it makes it inclear what you are using afterwards
import  graph_tool as gt

class Student:
  def __init__(self, name):
    self.name = name

class Professor:
    def __init__(self, name):
         self.name = name

g = gt.Graph()
s = Student("Luke")
p = Professor("Adam")
print(s.name)
print(p.name)

Luke
Adam

you create a property map with for the type of your users

g.vertex_properties['user_type'] = g.new_vertex_property('string')

This will be a property for your nodes. You can create how many you want and then set their user_type For exemple :

# creating a graph of 3 users
g.add_vertex(3)

# Let's set the first vertex as a professor
g.vp['user_type'][g.vertex(0)] = 'professor'

# and the subsequent as students
for v in g.get_vertices()[1:]:
    g.vp['user_type'][v] = 'student'

# let's add an edge from the professor to the first student
g.add_edge(g.vertex(0), g.vertex(1))

<Edge object with source '0' and target '1' at 0x7f2c30370c30>

Now you can draw your graph using the property

def translate_elts_to_gtshape(g, source_prop='user_type', tdict=None):
    """This function adds a new property map called 'shape' to graph g.  It is populated by translating the source_prop with the tdict (translation dictionary) which should map source_property entries to shapes"""
    if tdict is None:
        tdict = {'professor': 'square',
                 'student': 'circle'}

    # source_property map array.  It's of string type so we use get_2d_array
    svp = g.vp[source_prop].get_2d_array([0])[0]
    for k, v in tdict.items():
        # for each key, value pair of tdict we change the source value (ie k) to the shape (v)
        svp[svp == k] = v

    # and now we create the new property shape
    g.vertex_properties['shape'] = g.new_vertex_property('string')
    for v, p in zip(g.vertices(), svp):
        # and copy in it the shape
        g.vp['shape'][v] = p

    # we return the new property created, although it is already added to the graph as an internal property
    return g.vp['shape']

translate_elts_to_gtshape(g)

<PropertyMap object with key type 'Vertex' and value type 'string', for Graph 0x7f2c5c3d6510, at 0x7f2c2997c950>

# let's just check what we have in our newly created shape property
[g.vp['shape'][v] for v in g.vertices()]

['square', 'circle', 'circle']

# and now let's use it to draw our graph
import graph_tool.draw as gtd

pos = gtd.sfdp_layout(g)
gtd.graph_draw(g, pos=pos,
               output_size = (300, 300),
               output_file = 'mynetwork.png',   # where your file goes
               vertex_shape=g.vp['shape'],
               vertex_text=g.vertex_index)

<PropertyMap object with key type 'Vertex' and value type 'vector<double>', for Graph 0x7f2c5c3d6510, at 0x7f2c14291590>

Graph of objects using graph-tool

1 Answers1