Universal Spectroscopy and Imaging Data (USID)


If your data has a clear N-dimensional form, consider using NSID and pyNSID as simpler and more user-friendly alternatives to USID and pyUSID.


USID is an open, community-driven, self-describing, and standardized schema for representing imaging and spectroscopy data of any size, dimensionality, precision, instrument of origin, or modality. While the abstract USID schema can in theory be paired with any file format, it performs best within Hierarchical Data Format Files (HDF5) which have become the de-facto standard in scientific research. The combination of USID within HDF5 files is refered to as h5USID.

The flexible and open nature of both HDF5 and USID elevates USID data stored in HDF5 files to be curation-ready and therefore both meet the guidelines for data sharing issued to federally funded agencies and satisfy the implementation of digital data management as outlined by the United States Department of Energy. Traceability and reproducibility of both the experiments as well as the data processing routines in h5USID are captured through comprehensive and consistent metadata stored as attributes where appropriate, in addition to links created between the data source and the analysis results within the HDF5 file. Despite the fact that it was developed independently, h5USID already satisfies most of the requirements of FAIR principles and we intend to make necessary changes to satisfy all the recommendations put forward in the FAIR principles.

Please see our Arxiv paper for more information on the USID ecosystem.

Using h5USID files

  • pyUSID provides a convenient python interface to read and write to h5USID files.

More information

We encourage you to go through the motivation for USID, read more about USID, and see why it is best paired with HDF5. We have a few hands-on to illustrate how a few common data types can be represented in USID. We also provide detailed specifications on how the data and metadata are structured in h5USID files. We also have some guidelines on capturing contextual metadata to support the data.

Should you have any questions, please feel free to get in touch with us