Small bugfix for run_id readout and check for particle type.

orcasong/data_to_images.py

@@ -294,9 +294,13 @@ def get_file_particle_type(event_pump):
         file_particle_type = 'undefined'
     else:
         particle_type = event_pump[0]['McTracks'].type
-        if np.abs(particle_type) == 13:
+
+        # if mupage file: first mc_track is an empty neutrino track, second track is the first muon
+        if particle_type[0] == 0 and np.abs(particle_type[1]) == 13:
             file_particle_type = 'muon'
-        elif np.abs(particle_type) in [12, 14, 16]:
+
+        # the first mc_track is the primary neutrino if the input file is containing neutrino events
+        elif np.abs(particle_type[0]) in [12, 14, 16]:
             file_particle_type = 'neutrino'
         else:
             raise ValueError('The type of the particles in the "McTracks" folder, <', str(particle_type), '> is not known.')

orcasong/file_to_hits.py

@@ -132,10 +132,19 @@ def get_tracks(event_blob, file_particle_type, event_hits, prod_ident):
     """
     # parse EventInfo and Header information
     event_id = event_blob['EventInfo'].event_id[0]
-    if 'Header' in event_blob: # if Header exists in file, take run_id from it. Else take it from RawHeader.
+
+    # the run_id info in the EventInfo group is broken for ORCA neutrino and mupage files
+    # The Header run_id is the most reliable one.
+
+    if 'Header' in event_blob: # if Header exists in file, take run_id from it.
         run_id = event_blob['Header'].start_run.run_id.astype('float32')
     else:
-        run_id = event_blob['RawHeader'][0][0].astype('float32')
+        if file_particle_type == 'muon':
+            run_id = event_blob['RawHeader'][1][0].astype('float32')
+        elif file_particle_type == 'undefined': # currently used with random_noise files
+            run_id = event_blob['EventInfo'].run_id
+        else:
+            run_id = event_blob['RawHeader'][0][0].astype('float32')
 
     # collect all event_track information, dependent on file_particle_type