Installation

The bare installation of PHOTONAI Graph will get you up and running with machine learning graph analysis.

Info

Some analysis may require additional packages. PHOTONAI Graph will inform you about these requirements if needed. If you want to install them in advance please read additional packages.

Prerequisites

We recommend using anaconda environments. Alternatively you can simply use virtual environments or install PHOTONAI Graph system wide.

PHOTONAI Graph is currently tested with python 3.9

Bash

conda create -n photonai_graph
conda activate photonai_graph
conda install python=3.9

Installation via PIP

The PHOTONAI Graph library can simply be installed via pip:

Bash

pip install photonai-graph

If you prefer to use the dev version of PHOTONAI Graph:

Warning

This Version could be unstable due to active development

Bash

pip install git+https://github.com/wwu-mmll/photonai_graph@dev

Installation verification

You can verify your installation by starting python and runnning:

Python

import photonai_graph

If this command does not raise an error, you are ready to go.

Additional packages

To be able to use all modules of the toolbox you have to install these additional packages:

Bash

pip install tensorflow
pip install torch
pip install dgl
pip install grakel