def test_save_load_hspy(self, save_path_hdf5):
s = load(EMSOFT_FILE)
axes_manager = setup_axes_manager(["energy", "height", "width"])
assert isinstance(s, EBSDMasterPattern)
assert s.axes_manager.as_dictionary() == axes_manager
assert_dictionary(s.metadata.as_dictionary(), METADATA)
s.save(save_path_hdf5)
s2 = hs_load(save_path_hdf5, signal_type="EBSDMasterPattern")
assert isinstance(s2, EBSDMasterPattern)
assert s2.axes_manager.as_dictionary() == axes_manager
assert_dictionary(s2.metadata.as_dictionary(), METADATA)
s3 = hs_load(save_path_hdf5)
assert isinstance(s3, Signal2D)
s3.set_signal_type("EBSDMasterPattern")
assert isinstance(s3, EBSDMasterPattern)
assert s3.axes_manager.as_dictionary() == axes_manager
assert_dictionary(s.metadata.as_dictionary(), METADATA)
def test_original_metadata_save_load_cycle(self, save_path_hdf5):
s = nickel_ebsd_master_pattern_small()
omd_dict_keys = s.original_metadata.as_dictionary().keys()
desired_keys = [
"BetheList",
"EBSDMasterNameList",
"MCCLNameList",
"AtomData",
"Atomtypes",
"CrystalSystem",
"LatticeParameters",
"Natomtypes",
]
assert [k in omd_dict_keys for k in desired_keys]
s.save(save_path_hdf5)
s2 = hs_load(save_path_hdf5, signal_type="EBSDMasterPattern")
assert isinstance(s2, EBSDMasterPattern)
omd_dict_keys2 = s2.original_metadata.as_dictionary().keys()
assert [k in omd_dict_keys2 for k in desired_keys]
def test_load_save_hyperspy_cycle(self, tmp_path):
s = load(KIKUCHIPY_FILE)
# Perform decomposition to tests if learning results are
# maintained after saving, reloading and using set_signal_type
s.change_dtype(np.float32)
s.decomposition()
# Write both patterns and learning results to the HSpy file
# format
# file = tmp_path / "patterns.hspy"
os.chdir(tmp_path)
file = "patterns.hspy"
s.save(file)
# Reload data and use HyperSpy's set_signal_type function
s_reload = hs_load(file)
s_reload.set_signal_type("EBSD")
# Check signal type, patterns and learning results
assert isinstance(s_reload, EBSD)
assert np.allclose(s.data, s_reload.data)
assert np.allclose(s.learning_results.factors,
s_reload.learning_results.factors)