nomad-polymerization-reactions

A NOMAD plugin providing schemas, search apps, and data transformation utilities. The plugin has been primarily developed for hosting polymerization datasets extracted from the literature in a NOMAD deployment.

Availability

The plugin is available on the central NOMAD deployment. You can use the polymerization search app to access entries from publicly available datasets. These entries are based on the PolymerizationReaction schema defined in this plugin.

Associated Work

The dataset curated using this plugin has been used in the copolymer-reactivity project to train machine-learning models for classifying copolymerization types.

Installation

Install the package in your local environment with pip:

pip install nomad-polymerization-reactions

Adding this plugin to NOMAD Oasis

The plugin can be added to a NOMAD Oasis instance in a few steps, provided that you have access to the repository hosting the Oasis.

Read the NOMAD plugin documentation for all details on how to deploy the plugin on your NOMAD Oasis instance. If you want to get started with a new Oasis, read the installation guide.

Using the CLI for data transformation

The package provides utility functions for transforming JSON files into NOMAD entry archives. These archives have archive.json ending and are processed by NOMAD once uploaded. The m_def key in the archives helps NOMAD to identify which data schema to use (defined in schema_packages / polymerization.py).

Here's how you can convert a JSON file(s) into NOMAD entry archive(s) using the CLI command nomad-polymerization archive:

# create an archive with `Monomer` schema
nomad-polymerization archive monomer.json

# create an archive with `PolymerizationReaction` schema
nomad-polymerization archive polymerization_reaction.json --mode polymerization

# provide multiple filepaths or directory paths to convert multiple JSONs
nomad-polymerization archive /folder/poly1.json /folder/poly2.json  --mode polymerization
nomad-polymerization archive /folder --mode polymerization

# By default, the command creates archives in the current working directory. To create archives
# in the same directory as the JSONs, use `same-dir` flag:
nomad-polymerization archive /folder/monomer.json --same-dir

The JSON files used for transformation should have a fixed format. You can find the data models for JSON files in models.py.

License

Apache 2.0 - see LICENSE.