Writing Data to an HDF5 File Using H5Py Results in an Empty File

Question

I'm working on converting a large database for storage in an HDF5 file. To get familiar with H5Py (version 3.2.1) and HDF5, I read the docs for H5Py and wrote a small script that stores random data in an HDF5 file, shown below.

import h5py
import numpy as np


def main():
    f = h5py.File('testFile.hdf5', 'w')
    simBigData = np.random.randint(50, size=(24, 6), dtype=np.int32)
    simSmallData = np.random.randint(50, size=(8, 6), dtype=np.int32)
    simOut = np.random.randint(50, size=(8, 6), dtype=np.int32)
    grp = f.create_group('testGroup')
    dsBigData = grp.create_dataset('bigData', data=simBigData)
    dsSmallData = grp.create_dataset('smallData', data=simSmallData)
    dsOut = grp.create_dataset('out', data=simOut)
    print('HDF5 Data')
    print(f['testGroup/bigData'])
    print(f['testGroup/smallData'])
    print(f['testGroup/out'])
    f.close()


if __name__ == '__main__':
    main()

When I run this script, I get the following output, so something has definitely been written, at least in memory.

HDF5 Data
<HDF5 dataset "bigData": shape (24, 6), type "<i4">
<HDF5 dataset "smallData": shape (8, 6), type "<i4">
<HDF5 dataset "out": shape (8, 6), type "<i4">

I get an HDF5 file in my directory of about 5kB. However, when I open it with HDFView (version 2.11), I get a blank file. Clicking on the Metadata tab at the bottom, I get the following, which appears to show nothing in the file:

/ (0)
    Group size = 0
    Number of attributes = 0

How do I get H5Py to actually write the file correctly?

Answer 1

Per the comments by @hpaulj, I investigated the different versions. The version of HDFView in the Ubuntu repository is so old that it isn't able to open the generated HDF5 file. Switching to h5dump, I was able to verify the structure of my file was written correctly.