LabGraph

LabGraph is a streaming framework built by the Facebook Reality Labs Research team at Facebook. LabGraph is also mentioned in this story. More information can be found here.

Quick Start

Method 1 - using PyPI (Recommended)

Prerequisites:

• Python3.6+ (Python 3.8 recommended)
• Mac (Big Sur, Monterey), Windows and Linux (CentOS 7, CentOS 8, Ubuntu 20.04; Python3.6/3.8)
• Based on PyPa, the following Linux systems are also supported: Fedora 32+, Mageia 8+, openSUSE 15.3+, Photon OS 4.0+ (3.0+ with updates), Ubuntu 20.04+

pip install labgraph

For more documentation on installation and trouble shooting, please use this link.

Method 2 - building from source code

Prerequisites:

• Buck (Watchman also recommended)
• Python3.6-Python3.10
• Windows only: Build Tools for Visual Studio 2019

cd labgraph
python setup.py install

Method 3 - using Docker

Prerequisites:

• Docker

Setup:

docker login
docker build -t IMAGE_NAME:VERSION .
docker images
docker run -it -d Image_ID
docker ps -a
docker exec -it CONTAINER_ID bash

Testing:

To make sure things are working you can run the example:

python -m labgraph.examples.simple_viz

You can also run the test suite as follows:

python -m pytest --pyargs labgraph

or

RUN export LC_ALL=C.UTF-8
RUN export LANG=en_US.utf-8
. test_script.sh

Now go to the index and documentation to learn more!

License: LabGraph is MIT licensed, as found in the LICENSE file.