Are multiple "PointData" datasets supported using VTKHDF "ImageData".

Support
1

The documentation for VTKHDF seems to suggest that the following code:

import h5py
import numpy as np

with h5py.File("wtf.hdf",'w') as hdffile:

        # write support data
        vtkhdf_group = hdffile.create_group("VTKHDF")
        vtkhdf_group.attrs.create("Version", [1, 0])

        vtkhdf_group.attrs.create("Type",np.string_("ImageData"))
        whole_extent = (0, 3, 0, 1, 0, 0)
        vtkhdf_group.attrs.create("WholeExtent", whole_extent)

        vtkhdf_group.attrs.create("Origin", (0.0, 0.0, 0.0))
        vtkhdf_group.attrs.create("Spacing", (1.0, 1.0, 1.0))
        vtkhdf_group.attrs.create("Direction", (1.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 1.0))

        # create the pointdata group and the dataset inside it
        field_data_group = vtkhdf_group.create_group("PointData")
        field_data_group.attrs.create('Scalars', ["D1","D2"],) ## Doesn't work
        dset1 = field_data_group.create_dataset('D1',dtype=np.uint8,shape=(2,4))

        dset1[:, :] = np.reshape([0.0,1.0,2.0,3.0,4.0,5.0,6.0,7.0],(2, 4))

        dset2 = field_data_group.create_dataset('D2',dtype=np.uint8,shape=(2,4))

        dset2[:, :] = np.reshape([7.0,6.0,5.0,4.0,3.0,2.0,1.0,0.0],(2, 4))

would generate a valid file with two PointData datasets, but when viewing is attempted in Paraview 5.1.1 it crashes. This is just to confirm I am reading the limited documentation correctly.

2

I did not look at your dataset yet.

Yes, those correspond to multiple point data fields. You can take a look athttps://gitlab.kitware.com/danlipsa/vtkxml-to-vtkhdf
as an example that converts VTK XML datasets to VTK HDF datasets for image data and unstructured grid.

There is a VTK test you can try:
ctest -R HDFReader

That test reads the following HDF image dataset you can inspect, maybe you figure out what is the issue:
build/ExternalData/Testing/Data/mandelbrot-vti.hdf

Dan