Quickstart: Graphing Les Misérables

This simple example from our quickstart walkthrough introduces the basics of PyGraphistry. We also have more advanced tutorials avaiable.

You can download this notebook to run it locally.

In [1]:
import pandas
import graphistry

graphistry.register(key='Email [email protected] to get your API key')
In [2]:
# Parse CSV using Pandas
links = pandas.read_csv('data/lesmiserables.csv')
In [3]:
# Let's have a peak at our data by printing the first three rows
source target value
0 Napoleon Myriel 1
1 Mlle.Baptistine Myriel 8
2 Mme.Magloire Myriel 10
In [4]:
# Plot graph using the source/target columns as source/destination of edges
plotter = graphistry.bind(source='source', destination='target')
In [5]:
# New graph adding the number of encounters to edge labels.
links['label'] = links.value.map(lambda v: 'Num. Encounters: %d' % v)
plotter = plotter.bind(edge_label='label')

Controling Node Size and Color

We are going to use Igraph to color nodes by community and size them using pagerank. To install igraph, use pip install python-igraph

In [6]:
# Convert our graph from Pandas to Igraph
import igraph
ig = plotter.pandas2igraph(links)
IGRAPH D--- 77 254 -- 
+ attr: __nodeid__ (v), label (e), value (e)
WARNING: "node" is unbound, automatically binding it to "__nodeid__".
In [7]:
# We create two node attributes for pagerank and community
ig.vs['pagerank'] = ig.pagerank()
ig.vs['community'] = ig.community_infomap().membership 
IGRAPH D--- 77 254 -- 
+ attr: __nodeid__ (v), community (v), pagerank (v), label (e), value (e)
In [8]:
# The plotter can plot Igraph directly
plotter.bind(point_color='community', point_size='pagerank').plot(ig)