Fixed issues in metainfo converting arrays to scalar values even if shape was...

Fixed issues in metainfo converting arrays to scalar values even if shape was set. Simplified the HOMO detection in DOSNormalizer.

gui/src/components/App.js

@@ -138,10 +138,10 @@ function BetaSnack() {
       className={classes.snack}
       message={<span style={{color: 'white'}}>
         You are using a beta version of NOMAD ({version.label}). {
-          version.usesBetaData ? 'This version is not using the official data. Everything you upload here, might get lost.': ''
+          version.usesBetaData ? 'This version is not using the official data. Everything you upload here, might get lost.' : ''
         } Click <a style={{color: 'white'}} href={version.officialUrl}>here for the official NOMAD version</a>.
       </span>}
-       action={[
+      action={[
         <IconButton key={0} color="inherit" onClick={() => setUnderstood(true)}>
           <UnderstoodIcon />
         </IconButton>

nomad/metainfo/metainfo.py

@@ -822,20 +822,21 @@ class MSection(metaclass=MObjectMeta):  # TODO find a way to make this a subclas
                     (quantity_def, value))
             value = value.to(quantity_def.unit).magnitude
 
-        if len(quantity_def.shape) > 0 and type(value) != np.ndarray:
-            try:
-                value = np.asarray(value)
-            except TypeError:
-                raise TypeError(
-                    'Could not convert value %s of %s to a numpy array' %
-                    (value, quantity_def))
-        elif type(value) != quantity_def.type.type:
-            try:
-                value = quantity_def.type.type(value)
-            except TypeError:
-                raise TypeError(
-                    'Could not convert value %s of %s to a numpy scalar' %
-                    (value, quantity_def))
+        if type(value) != np.ndarray:
+            if len(quantity_def.shape) > 0:
+                try:
+                    value = np.asarray(value)
+                except TypeError:
+                    raise TypeError(
+                        'Could not convert value %s of %s to a numpy array' %
+                        (value, quantity_def))
+            elif type(value) != quantity_def.type.type:
+                try:
+                    value = quantity_def.type.type(value)
+                except TypeError:
+                    raise TypeError(
+                        'Could not convert value %s of %s to a numpy scalar' %
+                        (value, quantity_def))
 
         return self.__check_np(quantity_def, value)
 

nomad/normalizing/dos.py

@@ -13,9 +13,8 @@
 # limitations under the License.
 
 import numpy as np
-import bisect
 
-from nomad import config, units
+from nomad import config
 from nomad_dos_fingerprints import DOSFingerprint
 from nomad.datamodel.metainfo.public import section_dos_fingerprint
 from nomad.atomutils import get_volume
@@ -47,7 +46,8 @@ class DosNormalizer(Normalizer):
             for dos in section_dos:
                 dos_values = dos.dos_values
                 dos_energies = dos.dos_energies
-                if dos_energies is None or dos_values is None:
+                energy_reference_fermi = scc.energy_reference_fermi
+                if dos_energies is None or dos_values is None or energy_reference_fermi is None:
                     continue
 
                 # Normalize DOS values to be 1/J/atom/m^3
@@ -67,64 +67,59 @@ class DosNormalizer(Normalizer):
                 unit_cell_volume = get_volume(lattice_vectors.magnitude)
                 dos_values_normalized = dos_values / (number_of_atoms * unit_cell_volume)
 
-                # Normalize energies so that they are normalized to HOMO when
-                # it can be reliably detected and to Fermi level if HOMO could
-                # not be detected.
-                energy_reference_fermi = scc.energy_reference_fermi[0]
-                fermi_idx = (np.abs(dos_energies - energy_reference_fermi)).argmin()
+                # Normalize energies so that they are normalized to HOMO.
+                i_channel = 0
+                fermi_energy = energy_reference_fermi[i_channel]
+                fermi_idx = (np.abs(dos_energies - fermi_energy)).argmin()
                 energy_threshold = config.normalize.band_structure_energy_tolerance
                 value_threshold = 1e-8  # The DOS value that is considered to be zero
                 homo_found = False
                 zero_found = False
-                i_homo = 0
-                energy_distance = 0
+                energy_reference = fermi_energy
 
                 # Walk through the energies in descencing direction to see if a
                 # gap is nearby (see energy_threshold). If gap found, continue
                 # until HOMO found
-                while not homo_found:
+                idx = fermi_idx
+                while True:
                     try:
-                        value = dos_values_normalized[0, fermi_idx + i_homo]
-                        energy_distance = energy_reference_fermi - dos_energies[fermi_idx + i_homo]
+                        value = dos_values_normalized[i_channel, idx]
+                        energy_distance = fermi_energy - dos_energies[idx]
                     except IndexError:
                         break
                     if energy_distance.magnitude > energy_threshold and not zero_found:
                         break
                     if value <= value_threshold:
                         zero_found = True
-                    if zero_found and value > 0:
+                    if zero_found and value > value_threshold:
+                        energy_reference = dos_energies[idx + 1]
                         homo_found = True
-                        i_homo += 1
                         break
-                    i_homo -= 1
+                    idx -= 1
                 # If gap was not found in descending direction, check the
                 # ascending direction for a nearby (see energy_threshold) HOMO
                 # value
                 if not homo_found:
-                    i_homo = 1
-                    while not homo_found:
+                    idx = fermi_idx + 1
+                    while True:
                         try:
-                            value = dos_values_normalized[0, fermi_idx + i_homo]
-                            energy_distance = dos_energies[fermi_idx + i_homo] - energy_reference_fermi
+                            value = dos_values_normalized[i_channel, idx]
+                            energy_distance = dos_energies[idx] - fermi_energy
                         except IndexError:
                             break
                         if energy_distance.magnitude > energy_threshold:
                             break
                         if value <= value_threshold:
-                            homo_found = True
+                            energy_reference = dos_energies[idx]
                             break
-                        i_homo += 1
-                if homo_found:
-                    energy_reference = dos_energies[fermi_idx + i_homo]
-                else:
-                    energy_reference = energy_reference_fermi
+                        idx += 1
+
                 dos_energies_normalized = dos_energies - energy_reference
 
                 # Data for DOS fingerprint
                 dos_fingerprint = None
                 try:
-                    dos_energies = dos.dos_energies_normalized
-                    dos_fingerprint = DOSFingerprint().calculate(np.array(dos_energies), dos_normed, n_atoms=number_of_atoms)
+                    dos_fingerprint = DOSFingerprint().calculate(dos_energies_normalized, dos_values_normalized, n_atoms=number_of_atoms)
                 except Exception as e:
                     self.logger.error('could not generate dos fingerprint', exc_info=e)
 

tests/normalizing/test_dos.py

@@ -27,7 +27,7 @@ from nomad_dos_fingerprints import DOSFingerprint
 
 def test_fingerprint(dos_unpolarized_vasp):
     # Check if DOS fingerprint was created
-    backend_dos_fingerprint = dos_unpolarized_vasp['section_dos_fingerprint'][0]
+    backend_dos_fingerprint = dos_unpolarized_vasp.get_value('section_dos_fingerprint', 0)
     dos_fingerprint = DOSFingerprint().from_dict(dict(
         bins=backend_dos_fingerprint.bins,
         indices=backend_dos_fingerprint.indices,
@@ -40,19 +40,19 @@ def test_fingerprint(dos_unpolarized_vasp):
 
 
 # def test_dos_energies(dos_si_vasp: Backend, dos_si_exciting: Backend, dos_si_fhiaims: Backend):
-#   """For debugging.
-#   """
-#   x_exciting = dos_si_exciting.get_value('dos_energies_normalized', 0)
-#   y_exciting = dos_si_exciting.get_value('dos_values_normalized', 0)
-#   x_vasp = dos_si_vasp.get_value('dos_energies_normalized', 0)
-#   y_vasp = dos_si_vasp.get_value('dos_values_normalized', 0)
-#   x_fhiaims = dos_si_fhiaims.get_value('dos_energies_normalized', 0)
-#   y_fhiaims = dos_si_fhiaims.get_value('dos_values_normalized', 0)
-#   mpl.plot(x_vasp, y_vasp[0], label="VASP")
-#   mpl.plot(x_exciting, y_exciting[0], label="exciting")
-#   mpl.plot(x_fhiaims, y_fhiaims[0], label="FHI-aims")
-#   mpl.legend()
-#   mpl.show()
+#  """For debugging.
+#  """
+#  x_exciting = dos_si_exciting.get_value('dos_energies_normalized', 0)
+#  y_exciting = dos_si_exciting.get_value('dos_values_normalized', 0)
+#  x_vasp = dos_si_vasp.get_value('dos_energies_normalized', 0)
+#  y_vasp = dos_si_vasp.get_value('dos_values_normalized', 0)
+#  x_fhiaims = dos_si_fhiaims.get_value('dos_energies_normalized', 0)
+#  y_fhiaims = dos_si_fhiaims.get_value('dos_values_normalized', 0)
+#  mpl.plot(x_vasp, y_vasp[0], label="VASP")
+#  mpl.plot(x_exciting, y_exciting[0], label="exciting")
+#  mpl.plot(x_fhiaims, y_fhiaims[0], label="FHI-aims")
+#  mpl.legend()
+#  mpl.show()
 
 
 def test_dos_magnitude(dos_si_vasp: Backend, dos_si_exciting: Backend, dos_si_fhiaims: Backend):