Some code improvements for better performance (#125)

* Minor fixes

* Minor improvements for code style & performance

* Remove useless comments

* bump plugin versions

---------

Co-authored-by: Henning Schulze Eißing <h_schu55@uni-muenster.de>

fuse/plugins/detector_physics/s1_photon_hits.py

@@ -159,4 +159,4 @@ def get_n_photons(self, n_photons, positions):
 
         n_photon_hits = self.rng.binomial(n=n_photons, p=ly)
 
-        return n_photon_hits
+        return n_photon_hits

fuse/plugins/detector_physics/s1_photon_propagation.py

@@ -141,7 +141,7 @@ def setup(self):
                                )
 
         self.pmt_mask = np.array(self.gains) > 0  # Converted from to pe (from cmt by default)
-        self.turned_off_pmts = np.arange(len(self.gains))[np.array(self.gains) == 0]
+        self.turned_off_pmts = np.nonzero(np.array(self.gains) == 0)[0]
 
         self.spe_scaling_factor_distributions = init_spe_scaling_factor_distributions(self.photon_area_distribution)
 
@@ -307,7 +307,6 @@ def photon_timings(self,
         :param local_field: local field in the point of the deposit, 1d array of floats
         returns photon timing array"""
         _photon_timings = np.repeat(t, n_photon_hits)
-        _n_hits_total = len(_photon_timings)
 
         z_positions = np.repeat(positions[:, 2], n_photon_hits)
 

fuse/plugins/detector_physics/s2_photon_propagation.py

@@ -294,7 +294,7 @@ def setup(self):
                                )
 
         self.pmt_mask = np.array(self.gains) > 0  # Converted from to pe (from cmt by default)
-        self.turned_off_pmts = np.arange(len(self.gains))[np.array(self.gains) == 0]
+        self.turned_off_pmts = np.nonzero(np.array(self.gains) == 0)[0]
 
         self.spe_scaling_factor_distributions = init_spe_scaling_factor_distributions(self.photon_area_distribution)
 
@@ -450,8 +450,7 @@ def photon_channels(self, n_electron, z_obs, positions, drift_time_mean, n_photo
 
         channels = np.arange(self.n_tpc_pmts).astype(np.int64)
         top_index = np.arange(self.n_top_pmts)
-        channels_bottom = np.arange(self.n_top_pmts, self.n_tpc_pmts)
-        bottom_index = np.array(channels_bottom)
+        bottom_index = np.arange(self.n_top_pmts, self.n_tpc_pmts)
 
         if self.diffusion_constant_transverse > 0:
             pattern = self.s2_pattern_map_diffuse(n_electron, z_obs, positions, drift_time_mean)  # [position, pmt]

fuse/plugins/detector_physics/secondary_scintillation.py

@@ -15,7 +15,7 @@
 @export
 class SecondaryScintillation(FuseBasePlugin):
 
-    __version__ = "0.1.3"
+    __version__ = "0.1.4"
 
     depends_on = ("drifted_electrons","extracted_electrons" ,"electron_time")
     provides = ("s2_photons", "s2_photons_sum")
@@ -188,9 +188,10 @@ def compute(self, interactions_in_roi, individual_electrons):
         if self.s2_gain_spread:
             n_photons_per_ele += self.rng.normal(0, self.s2_gain_spread, len(n_photons_per_ele)).astype(np.int64)
 
-        sum_photons_per_interaction = [np.sum(x) for x in np.split(n_photons_per_ele, np.cumsum(interactions_in_roi[mask]["n_electron_extracted"]))[:-1]]
+        electron_indices = np.cumsum(interactions_in_roi[mask]["n_electron_extracted"])
+        sum_photons_per_interaction = np.add.reduceat(n_photons_per_ele, np.r_[0, electron_indices[:-1]])
 
-        n_photons_per_ele[n_photons_per_ele < 0] = 0
+        n_photons_per_ele = np.clip(n_photons_per_ele, 0, None)
 
         reorder_electrons = np.argsort(individual_electrons, order = ["order_index", "time"])
 

fuse/plugins/micro_physics/find_cluster.py

@@ -16,7 +16,7 @@
 @export
 class FindCluster(FuseBasePlugin):
 
-    __version__ = "0.1.2"
+    __version__ = "0.1.3"
 
     depends_on = ("geant4_interactions")
 
@@ -72,13 +72,13 @@ def find_cluster(interactions, cluster_size_space, cluster_size_time):
         time_cluster = simple_1d_clustering(interactions["time"], cluster_size_time)
 
         # Splitting into time cluster and apply space clustering space:
-        cluster_id = np.zeros(len(interactions), dtype=np.int32)
         spacial_cluster = np.zeros(len(interactions), dtype=np.int32)
 
         _t_clusters = np.unique(time_cluster)
         for _t in _t_clusters:
-            _cl = _find_cluster(interactions[time_cluster == _t], cluster_size_space=cluster_size_space)
-            spacial_cluster[time_cluster == _t] = _cl
+            time_cluster_mask = time_cluster == _t
+            _cl = _find_cluster(interactions[time_cluster_mask], cluster_size_space=cluster_size_space)
+            spacial_cluster[time_cluster_mask] = _cl
         _, cluster_id = np.unique((time_cluster, spacial_cluster), axis=1, return_inverse=True)
 
         return cluster_id
@@ -95,8 +95,8 @@ def _find_cluster(x, cluster_size_space):
     """
     db_cluster = DBSCAN(eps=cluster_size_space, min_samples=1)
 
-    #Conversion from numpy structured array to regular array 
-    xprime = np.array(x[['x', 'y', 'z']].tolist())
+    # Conversion from numpy structured array to regular array
+    xprime = np.stack((x['x'], x['y'], x['z']), axis=1)
 
     return db_cluster.fit_predict(xprime)