I want to read the three dimensional data from a h5py as a numpy array.
61.42836 61.42836 8.47704
When reading, use the index slicing provided by h5py, which is a great feature. This line
with h5py.File('File') as f: data = np.array(f['group/data']) #reads the whole (some GB large) dataset at once. However, #if you do data = f['group/data']
then you can ‘treat’ data as a numpy array (although it is an h5py object). For example, data would yield the 4th sample (and read only this piece of data), whereas data[:, 5] would read and return the full trajectory of particle.