Repolode
Submit
graphistry

graphistry/pygraphistry

Python library for loading, shaping, embedding and visually exploring large graphs, with optional GPU acceleration via RAPIDS and the Graphistry visual analyzer.

Open-source alternative to:Gephi

View on GitHub
Stars
2,577
Forks
232
Open issues
336
Last commit
6 days ago

What is it?

PyGraphistry is a Python library for turning tabular data into graphs and exploring them interactively. It works with pandas DataFrames and, in GPU mode, cuDF, so data scientists can prepare edge and node tables in familiar code and send them to a Graphistry visualization server for interactive exploration with filtering, drilldowns and timebars.

The library includes GFQL, a dataframe-native graph query language with Cypher-style syntax that runs on the current graph or remotely. It also provides built-in graph analytics such as PageRank through igraph or cuGraph, UMAP embeddings, and graph ML helpers under the optional graphistry[ai] extra.

PyGraphistry offers connectors for databases and platforms including Neo4j, Neptune, TigerGraph, Splunk, PostgreSQL and Databricks, and interoperates with NetworkX and Apache Arrow. Visualizations can be embedded through iframes or customized via JavaScript and REST APIs.

Key features

  • 01Loads pandas and cuDF dataframes as graph edges and nodes with simple bindings.
  • 02Runs graph queries with GFQL using Cypher-style syntax on the current graph or remotely.
  • 03Computes analytics such as PageRank via igraph or cuGraph and produces UMAP embeddings.
  • 04Connects to graph databases and platforms including Neo4j, Neptune, TigerGraph and Splunk.
  • 05Renders interactive visualizations with filtering, drilldowns and timebars, embeddable via iframe or customizable via JavaScript and REST.

Use cases

  • Exploring relationships in fraud, cybersecurity logs or transaction data as interactive graphs.
  • Visualizing social networks, protein interactions or other large edge lists with millions of edges.
  • Running graph analytics and embeddings on tabular data in a Jupyter notebook.
  • Querying graph data with Cypher-style patterns from a Python workflow without a dedicated graph database.

Quick start

  1. 1

    Install the minimal core package with pip

    pip install graphistry
  2. 2

    Load a DataFrame and bind it as a graph with graphistry.edges(df, 'src', 'dst')

  3. 3

    Register with a Graphistry server account, then call plot() to view the graph

Things to know

  • Interactive plots upload to a Graphistry server, so you need an account on the hub or a self-hosted server.
  • GPU acceleration is optional but requires RAPIDS and an NVIDIA GPU for the local GPU mode.

In collections

Run and train AI on GPUs

Topics

#graph-visualization#graph-analytics#gpu#network-analysis#pandas#rapids#jupyter#graph-query-language#csv#cudf

FAQ

What is pygraphistry?
Python library for loading, shaping, embedding and visually exploring large graphs, with optional GPU acceleration via RAPIDS and the Graphistry visual analyzer. PyGraphistry is a Python library for turning tabular data into graphs and exploring them interactively. It works with pandas DataFrames and, in GPU mode, cuDF, so data scientists can prepare edge and node tables in familiar code and send them to a Graphistry visualization server for interactive exploration with filtering, drilldowns and timebars.
Is pygraphistry free to use? What license does it have?
pygraphistry is open source under the BSD-3-Clause license. Check the license terms before using it in commercial products.
Is pygraphistry still maintained?
Actively maintained. The last commit to pygraphistry was on Nov 2, 2026.
What can I use pygraphistry for?
Exploring relationships in fraud, cybersecurity logs or transaction data as interactive graphs. Visualizing social networks, protein interactions or other large edge lists with millions of edges. Running graph analytics and embeddings on tabular data in a Jupyter notebook. Querying graph data with Cypher-style patterns from a Python workflow without a dedicated graph database.
How do I install pygraphistry?
Install the minimal core package with pip: pip install graphistry Load a DataFrame and bind it as a graph with graphistry.edges(df, 'src', 'dst') Register with a Graphistry server account, then call plot() to view the graph
Where can I get pygraphistry?
The source code and installation instructions are on GitHub: https://github.com/graphistry/pygraphistry

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects