irdf

Jupyter extension for interacting with RDF graphs


Keywords
ipython, jupyter, visualization, interaction, interact
License
MIT
Install
pip install irdf==1.4

Documentation

irdf: Interactive RDF graph widgets in Jupyter

This package allows you to visualize your RDFlib graph objects in two ways: as triple tables or a network visualization.

In the triple tables, you can filter the triples by subject, predicate or object value.

In the network visualizations, you can see the relations between entities, and the literals associated with them.

Install

pip install irdf

Usage

import rdflib
import irdf

# Create a Graph
g = rdflib.Graph()

# Parse in an RDF file hosted on the Internet
g.parse('http://dbpedia.org/data/Vrije_Universiteit_Amsterdam.ttl')
# Get an interactive table
irdf.table(g)

table

# Get an interactive graph (filtering on certain properties)
# Click on nodes to see associated literals!
irdf.graph(g, p_filter=['/ontology/', 'depiction'], lang='en')

graph

Wikidata

In Wikidata, the labels of properties are associated with special property entities. To get nice labels on edges in your visualization, you can specify a custom SPARQL property path to fetch them:

import rdflib, irdf
g = rdflib.Graph()
g.parse('https://www.wikidata.org/wiki/Special:EntityData/Q1065414.ttl')
label_path = '(^wikibase:directClaim)*/rdfs:label'
irdf.graph(g, p_filter=['/direct/'], lang='en', label_path=label_path)

Acknowledgements

This library is based on the great work of ipycytoscape and itables.