HDF5 is a community standard file format that is fast and handles arbitrarily large datasets. HDF5 may be thought of as a filesystem withing a file, with access from most popular data science languages, including C, C++, Fortran, Julia Matlab, Python, R, etc. NetCDF4 is built upon HDF5 and is popular in the geoscience community.
Here are posts we’ve written discussing use of the HDF5 library.
- 2020-11-25 – Symbolic soft links in HDF5 / NetCDF4
- 2020-11-17 – Linking HDF5 with CMake for C, C++ and Fortran
- 2020-09-27 – Python ImportError vs. ModuleNotFoundError
- 2020-07-15 – Easy HDF5 Fortran interface
- 2020-07-13 – Easy HDF5 / NetCDF4 Matlab interface
- 2020-06-07 – Compare HDF5 data values
- 2020-03-21 – NetCDF4 segfault on file open
- 2020-02-17 – CMake FindHDF5 improvements
- 2020-01-10 – HDF5 on Intel oneAPI for Windows
- 2017-12-14 – NetCDF4 vs. HDF5 for large datasets
- 2017-03-29 – Testing h5py boolean vs uint8 write speed
- 2016-04-23 – Fix Python HDF5 library version mismatch