nrrd to nifti

nrrd_to_nifti.py

import nrrd # pip install pynrrd
import nibabel as nib # pip install nibabel
import numpy as np

# load nrrd 
_nrrd = nrrd.read('/path/to/nrrd.nrrd')
data = _nrrd[0]
header = _nrrd[1]
print data.shape, header

# save nifti
img = nib.Nifti1Image(data, np.eye(4))
nib.save(img,'/path/to/nifti.nii.gz')

This code is elegant in its simplicity and potentially really useful. My main concern is you set the SForm to an identity matrix, which does not preserve the spatial information (image orientation, voxel size). Can your script be updated to retain this information? The combination of space origin, space and space directions fields in NRRD can describe this transform.

For Python3 compatibility you might want to change print data.shape, header to read

print(data.shape)
print(header)