How to Use NOMAD's Workflow Utilities to Improve Data Management and Facilitate Discovery in Materials Science¶
NOMAD [nomad-lab.eu] [1] is an open-source, community-driven data infrastructure that supports automated (meta)data extraction from a wide range of simulations, including ab initio and advanced many-body calculations, as well as molecular dynamics simulations. NOMAD also provides extensive customization capabilities to support experimental data. NOMAD allows users to store both standardized and custom complex workflows, which streamline data provenance storage and analysis, facilitating efficient curation of AI-ready datasets. This tutorial series will focus on recently developed workflow functionalities and utilities within the NOMAD infrastructure, with a step-by-step guide for storing a custom project workflow that contains tasks involving a variety of distinct data sources. Attendees can then use this knowledge to transform their day-to-day project data management, or even to interface with the NOMAD repository in a high-throughput fashion, opening improved discovery pipelines by leveraging the benefits of NOMAD’s comprehensive and FAIR-compliant data management system [2]. Attendees are welcome to simply watch the demonstration or to follow along on their laptops. For the latter, preparation instructions will be provided in advance below.
[1] Scheidgen, M. et al., JOSS 8, 5388 (2023).
[2] Scheffler, M. et al., Nature 604, 635-642 (2022).
Curious about more NOMAD features?¶
You can find several presentations, as well as an information booth, here at the DPG Spring Meeting of the Condensed Matter Section (SKM) 2025. For times and locations, click here.
Tutorial preparation¶
1. Create a NOMAD account at NOMAD Central Deployment¶
Click LOGIN/REGISTER at the top right.
2. Install the nomad-utility-workflows module¶
Open a terminal and create a virtual environment with python==3.11 (It may be possible to use python>=3.9, but the module has only been fully tested with 3.11):
Install missing Python 3.11 interpreter
To install Python 3.11 interpreter:
Download the installer from the official Python website and run it. Make sure to check the box that says "Add Python 3.11 to PATH" during installation.
Activate the Python virtual environment:
Upgrade pip and install uv (recommended):
Install the latest pypi version of the plugin using pip:
uv pip install "nomad-utility-workflows>=0.0.19" --index-url https://gitlab.mpcdf.mpg.de/api/v4/projects/2187/packages/pypi/simple
Install python-dotenv package.
In order to use a Jupyter notebook in the following, install ipython and then create a Jupyter kernel for this venv (this kernel can then be be identified and loaded into your IDE):
uv pip install python-dotenv
uv pip install --upgrade ipython
uv pip install --upgrade ipykernel
ipython kernel install --user --name=DPG-tutorial
Now you should be able to simply launch a Jupyter notebook browser with jupyter notebook in the terminal.
Open a (new) .ipynb file, and then select DPG-tutorial from the kernel list.