Propnet is a knowledge graph for materials science. Given a set of properties about a given material, it can try to derive additional properties of that material using models. It is intended to integrate closely with the Materials Project and its large database of materials properties. However, it is designed to be agnostic to the data source, with support for other open-source databases planned.
Propnet can also be used as a library, each model can be imported and used directly in your own code, giving you tested, unit-aware implementations of common materials science models.
See propnet.lbl.gov for a web app based on the package.
For usage information, it is currently best to view our example Binder notebook, linked below. We do have a documentation website under construction at propnet.readthedocs.io.
Hat tip to @jdagdelen for the Propnet name. This project is funded by Toyota Research Institute.