Reference

HDF5

HDF5 (Hierarchical Data Format) is a format for storing large, complex scientific datasets in a single file. It organizes multidimensional arrays and metadata in a hierarchy, and is widely used in research, engineering, and machine learning.

Files & formatsGeneral

HDF5

Also known as: .h5 file, .hdf5 file, Hierarchical Data Format, HDF

HDF5 (Hierarchical Data Format) is a format for storing large, complex scientific datasets in a single file. It organizes multidimensional arrays and metadata in a hierarchy, and is widely used in research, engineering, and machine learning.

  • Hierarchical format for large scientific datasets
  • Stores multidimensional arrays plus metadata
  • Binary; read via scientific libraries and viewers

What HDF5 stores

An HDF5 file (.h5 or .hdf5) acts like a self-contained file system: it holds datasets (typically large numeric arrays) organized in groups, each tagged with attributes and metadata.

It is built for big numerical data — simulation output, sensor logs, satellite imagery, and model weights — that would be impractical as flat text. The format supports compression and reading just the slices you need.

Files and tools

HDF5 is a binary format read through libraries in scientific computing environments rather than a text editor. Viewer tools let you browse the internal hierarchy.

Files can be very large because they hold the actual data arrays, but built-in compression and partial reads keep working with them efficient.

Related terms

Keep reading the reference.