Installation¶

Prerequisites¶

Python 3.11+
A graph database (ArangoDB, Neo4j, TigerGraph, FalkorDB, Memgraph, or NebulaGraph) if you plan to use database features

Installation Methods¶

Using pip¶

pip install graflo

Using uv (recommended)¶

uv add graflo

From Source¶

Clone the repository:

git clone https://github.com/growgraph/graflo.git
cd graflo

Install with development dependencies (pytest, ty, pre-commit):

uv sync --extra dev

To build the documentation locally, add the docs extra:

uv sync --extra dev --extra docs

Optional extras¶

The default package includes RDF/SPARQL support (rdflib, SPARQLWrapper) and graph database clients. Optional project.optional-dependencies extras are tooling only (dev, docs, plot)—they do not toggle ingestion features:

Extra	Purpose
`dev`	Development: `pytest`, `ty`, `pre-commit`
`docs`	Building this site: MkDocs and plugins
`plot`	`plot_manifest` / schema diagrams via `pygraphviz`

pip¶

pip install "graflo[dev]"
pip install "graflo[docs]"
pip install "graflo[plot]"
# combine as needed, e.g.:
pip install "graflo[dev,docs,plot]"

uv¶

From another project:

uv add graflo
uv add "graflo[plot]"  # optional: plot_manifest

From a clone of this repository:

uv sync --extra dev
uv sync --extra plot

`plot` extra (Graphviz)¶

Install the system Graphviz libraries first (e.g. Debian/Ubuntu: apt install graphviz graphviz-dev), then install graflo[plot].

Verifying Installation¶

To verify your installation, you can run:

import graflo
print(graflo.__version__)

Spinning up databases¶

Instructions on how to spin up ArangoDB, Neo4j, TigerGraph, FalkorDB, Memgraph, NebulaGraph, and Apache Fuseki as Docker containers using docker compose are provided here: github.com/growgraph/graflo/docker

Configuration¶

After installation, you may need to configure your graph database connection. See the Quick Start Guide for details on setting up your environment.

For more detailed troubleshooting, refer to the API Reference or open an issue on GitHub.