def characterize_coincidences(input_file: str, output_file: str, rmap: str):
"""
This function extracts the relevant information on position, time,
sensor response, kind (point-like or not)) of the
two gamma interactions of a coincidences.
"""
DataSiPM = db.DataSiPMsim_only('petalo', 0) # full body PET
DataSiPM_idx = DataSiPM.set_index('SensorID')
### parameters for single photoelectron convolution in SiPM response
tau_sipm = [100, 15000]
time_window = 5000
time = np.arange(0, time_window)
spe_resp, norm = shf.normalize_sipm_shaping(time, tau_sipm)
thr_r = 4 # threshold use to create R map (pe)
thr_phi = 4 # threshold on charge to reconstruct phi (pe)
thr_z = 4 # threshold on charge to reconstruct z (pe)
thr_e = 2 # threshold on charge (pe)
n_pe = 1 # number of first photoelectrons to be considered for timestamp
sigma_sipm = 40 #ps SiPM jitter
sigma_elec = 30 #ps electronic jitter
Rpos = load_map(rmap,
group="Radius",
node="f4pes150bins",
x_name="PhiRms",
y_name="Rpos",
u_name="RposUncertainty")
c0 = c1 = 0
true_r1, true_phi1, true_z1 = [], [], []
reco_r1, reco_phi1, reco_z1 = [], [], []
true_r2, true_phi2, true_z2 = [], [], []
reco_r2, reco_phi2, reco_z2 = [], [], []
sns_response1, sns_response2 = [], []
### PETsys thresholds to extract the timestamp
timestamp_thr = 0.25
first_sipm1 = []
first_sipm2 = []
first_time1 = []
first_time2 = []
true_time1, true_time2 = [], []
touched_sipms1, touched_sipms2 = [], []
photo1, photo2 = [], []
max_hit_distance1, max_hit_distance2 = [], []
event_ids = []
try:
sns_response = load_mcsns_response(input_file)
except ValueError:
print(f'File {input_file} not found')
exit()
except OSError:
print(f'File {input_file} not found')
exit()
except KeyError:
print(f'No object named MC/sns_response in file {input_file}')
exit()
print(f'Analyzing file {input_file}')
fluct_sns_response = snsf.apply_charge_fluctuation(sns_response,
DataSiPM_idx)
tof_bin_size = read_sensor_bin_width_from_conf(input_file, tof=True)
particles = load_mcparticles(input_file)
hits = load_mchits(input_file)
tof_response = load_mcTOFsns_response(input_file)
events = particles.event_id.unique()
charge_range = (
2000, 2250
) # range to select photopeak - to be adjusted to the specific case
print(f'Number of events in file = {len(events)}')
for evt in events:
evt_sns = fluct_sns_response[fluct_sns_response.event_id == evt]
evt_sns = rf.find_SiPMs_over_threshold(evt_sns, threshold=thr_e)
if len(evt_sns) == 0:
continue
ids_over_thr = evt_sns.sensor_id.astype('int64').values
evt_parts = particles[particles.event_id == evt]
evt_hits = hits[hits.event_id == evt]
evt_tof = tof_response[tof_response.event_id == evt]
evt_tof = evt_tof[evt_tof.sensor_id.isin(-ids_over_thr)]
if len(evt_tof) == 0:
continue
### If one wants to select only pure photoelectric events
### add the following lines
#select, true_pos = mcf.select_photoelectric(evt_parts, evt_hits)
#if not select: continue
#if (len(true_pos) == 1) & (evt_hits.energy.sum() > 0.511):
# continue
pos1, pos2, q1, q2, true_pos1, true_pos2, true_t1, true_t2, sns1, sns2 = rf.reconstruct_coincidences(
evt_sns, charge_range, DataSiPM_idx, evt_parts, evt_hits)
if len(pos1) == 0 or len(pos2) == 0:
c0 += 1
continue
q1 = np.array(q1)
q2 = np.array(q2)
pos1 = np.array(pos1)
pos2 = np.array(pos2)
r1, phi1, z1 = rf.reconstruct_position(q1, pos1, Rpos, thr_r, thr_phi,
thr_z)
r2, phi2, z2 = rf.reconstruct_position(q2, pos2, Rpos, thr_r, thr_phi,
thr_z)
if (r1 > 1.e8) or (r2 > 1.e8):
c1 += 1
continue
## Use absolute times in units of ps
times = evt_tof.time_bin.values * tof_bin_size / units.ps
## add SiPM jitter, if different from zero
if sigma_sipm > 0:
times = np.round(np.random.normal(times, sigma_sipm))
evt_tof.insert(len(evt_tof.columns), 'time',
times.astype(int)) # here we have bins of 1 ps
## produce a TOF dataframe with convolved time response
tof_sns = evt_tof.sensor_id.unique()
evt_tof_exp_dist = []
for s_id in tof_sns:
tdc_conv = shf.sipm_shaping_convolution(evt_tof, spe_resp, s_id,
time_window)
tdc_conv_df = shf.build_convoluted_df(evt, s_id, tdc_conv)
if sigma_elec > 0:
tdc_conv_df = tdc_conv_df.assign(
time=np.random.normal(tdc_conv_df.time.values, sigma_elec))
tdc_conv_df = tdc_conv_df[tdc_conv_df.charge > timestamp_thr /
norm]
tdc_conv_df = tdc_conv_df[tdc_conv_df.time ==
tdc_conv_df.time.min()]
evt_tof_exp_dist.append(tdc_conv_df)
evt_tof_exp_dist = pd.concat(evt_tof_exp_dist)
try:
min_id1, min_id2, min_t1, min_t2 = rf.find_coincidence_timestamps(
evt_tof_exp_dist, sns1, sns2, n_pe)
ave_pos1 = rf.calculate_average_SiPM_pos(min_id1, DataSiPM_idx)
ave_pos2 = rf.calculate_average_SiPM_pos(min_id2, DataSiPM_idx)
first_sipm1.append(ave_pos1)
first_sipm2.append(ave_pos2)
except WaveformEmptyTable:
print(f'TOF dataframe has no minimum time for event {evt}')
_, _, min_t1, min_t2 = [-1], [-1], -1, -1
first_sipm1.append(np.array([0, 0, 0]))
first_sipm2.append(np.array([0, 0, 0]))
first_time1.append(min_t1)
first_time2.append(min_t2)
## extract information about the interaction being photoelectric
phot, phot_pos = mcf.select_photoelectric(evt_parts, evt_hits)
if not phot:
phot1 = False
phot2 = False
else:
scalar_prod = true_pos1.dot(phot_pos[0])
if scalar_prod > 0:
phot1 = True
phot2 = False
else:
phot1 = False
phot2 = True
if len(phot_pos) == 2:
if scalar_prod > 0:
phot2 = True
else:
phot1 = True
## extract information about the interaction being photoelectric-like
distances1 = rf.find_hit_distances_from_true_pos(evt_hits, true_pos1)
max_dist1 = distances1.max()
distances2 = rf.find_hit_distances_from_true_pos(evt_hits, true_pos2)
max_dist2 = distances2.max()
event_ids.append(evt)
reco_r1.append(r1)
reco_phi1.append(phi1)
reco_z1.append(z1)
true_r1.append(np.sqrt(true_pos1[0]**2 + true_pos1[1]**2))
true_phi1.append(np.arctan2(true_pos1[1], true_pos1[0]))
true_z1.append(true_pos1[2])
sns_response1.append(sum(q1))
touched_sipms1.append(len(q1))
true_time1.append(true_t1 / units.ps)
photo1.append(phot1)
max_hit_distance1.append(max_dist1)
reco_r2.append(r2)
reco_phi2.append(phi2)
reco_z2.append(z2)
true_r2.append(np.sqrt(true_pos2[0]**2 + true_pos2[1]**2))
true_phi2.append(np.arctan2(true_pos2[1], true_pos2[0]))
true_z2.append(true_pos2[2])
sns_response2.append(sum(q2))
touched_sipms2.append(len(q2))
true_time2.append(true_t2 / units.ps)
photo2.append(phot2)
max_hit_distance2.append(max_dist2)
a_true_r1 = np.array(true_r1)
a_true_phi1 = np.array(true_phi1)
a_true_z1 = np.array(true_z1)
a_reco_r1 = np.array(reco_r1)
a_reco_phi1 = np.array(reco_phi1)
a_reco_z1 = np.array(reco_z1)
a_sns_response1 = np.array(sns_response1)
a_touched_sipms1 = np.array(touched_sipms1)
a_first_sipm1 = np.array(first_sipm1)
a_first_time1 = np.array(first_time1)
a_true_time1 = np.array(true_time1)
a_photo1 = np.array(photo1)
a_max_hit_distance1 = np.array(max_hit_distance1)
a_true_r2 = np.array(true_r2)
a_true_phi2 = np.array(true_phi2)
a_true_z2 = np.array(true_z2)
a_reco_r2 = np.array(reco_r2)
a_reco_phi2 = np.array(reco_phi2)
a_reco_z2 = np.array(reco_z2)
a_sns_response2 = np.array(sns_response2)
a_touched_sipms2 = np.array(touched_sipms2)
a_first_sipm2 = np.array(first_sipm2)
a_first_time2 = np.array(first_time2)
a_true_time2 = np.array(true_time2)
a_photo2 = np.array(photo2)
a_max_hit_distance2 = np.array(max_hit_distance2)
a_event_ids = np.array(event_ids)
np.savez(output_file,
a_true_r1=a_true_r1,
a_true_phi1=a_true_phi1,
a_true_z1=a_true_z1,
a_true_r2=a_true_r2,
a_true_phi2=a_true_phi2,
a_true_z2=a_true_z2,
a_reco_r1=a_reco_r1,
a_reco_phi1=a_reco_phi1,
a_reco_z1=a_reco_z1,
a_reco_r2=a_reco_r2,
a_reco_phi2=a_reco_phi2,
a_reco_z2=a_reco_z2,
a_touched_sipms1=a_touched_sipms1,
a_touched_sipms2=a_touched_sipms2,
a_sns_response1=a_sns_response1,
a_sns_response2=a_sns_response2,
a_first_sipm1=a_first_sipm1,
a_first_time1=a_first_time1,
a_first_sipm2=a_first_sipm2,
a_first_time2=a_first_time2,
a_true_time1=a_true_time1,
a_true_time2=a_true_time2,
a_photo1=a_photo1,
a_photo2=a_photo2,
a_max_hit_distance1=a_max_hit_distance1,
a_max_hit_distance2=a_max_hit_distance2,
a_event_ids=a_event_ids)
print(f'Not a coincidence: {c0}')
print(f'Not passing threshold to reconstruct position = {c1}')
file_name = file_full.format(ifile)
try:
sns_response = pd.read_hdf(file_name, 'MC/sns_response')
except ValueError:
print('File {} not found'.format(file_name))
continue
except OSError:
print('File {} not found'.format(file_name))
continue
except KeyError:
print('No object named MC/sns_response in file {0}'.format(file_name))
continue
print('Analyzing file {0}'.format(file_name))
sel_df = rf.find_SiPMs_over_threshold(sns_response, threshold)
particles = pd.read_hdf(file_name, 'MC/particles')
hits = pd.read_hdf(file_name, 'MC/hits')
events = particles.event_id.unique()
for evt in events[:]:
### Select photoelectric events only
evt_parts = particles[particles.event_id == evt]
evt_hits = hits[hits.event_id == evt]
select, true_pos = mcf.select_photoelectric(evt_parts, evt_hits)
if not select: continue
sns_resp = sel_df[sel_df.event_id == evt]
if len(sns_resp) == 0: continue
tof_bin_size = read_sensor_bin_width_from_conf(file_name)
particles = pd.read_hdf(file_name, 'MC/particles')
hits = pd.read_hdf(file_name, 'MC/hits')
tof_response = pd.read_hdf(file_name, 'MC/tof_sns_response')
events = particles.event_id.unique()
print(len(events))
#charge_range = (1000, 1500) # for the ring
charge_range = (1050, 1300) # for full body PET
for evt in events[:]:
evt_sns = sns_response[sns_response.event_id == evt]
evt_sns = rf.find_SiPMs_over_threshold(evt_sns, threshold=2)
if len(evt_sns) == 0:
continue
evt_parts = particles[particles.event_id == evt]
evt_hits = hits[hits.event_id == evt]
evt_tof = tof_response[tof_response.event_id == evt]
pos1, pos2, q1, q2, true_pos1, true_pos2, true_t1, true_t2, min_id1, min_id2, min_t1, min_t2 = rf.reconstruct_coincidences(
evt_sns, evt_tof, charge_range, DataSiPM_idx, evt_parts, evt_hits)
if len(pos1) == 0 or len(pos2) == 0:
c0 += 1
continue
q1 = np.array(q1)
q2 = np.array(q2)
for evt in events[:]:
### Select photoelectric events only
evt_parts = particles[particles.event_id == evt]
evt_hits = hits[hits.event_id == evt]
select, true_pos = mcf.select_photoelectric(evt_parts, evt_hits)
if not select: continue
if (len(true_pos) == 1) & (evt_hits.energy.sum() > 0.511):
continue
sns_evt = sns_response[sns_response.event_id == evt]
evt_tof = tof_response[tof_response.event_id == evt]
sns_resp_r = rf.find_SiPMs_over_threshold(sns_evt, threshold=thr_r)
sns_resp_phi = rf.find_SiPMs_over_threshold(sns_evt, threshold=thr_phi)
sns_resp_z = rf.find_SiPMs_over_threshold(sns_evt, threshold=thr_z)
sns_resp_e = rf.find_SiPMs_over_threshold(sns_evt, threshold=thr_e)
_, _, pos1, pos2, q1, q2 = rf.assign_sipms_to_gammas(
sns_resp_r, true_pos, DataSiPM_idx)
if len(q1) == len(q2) == 0:
c1 += 1
r1 = r2 = None
if len(pos1) > 0:
pos1_phi = rf.from_cartesian_to_cyl(np.array(pos1))[:, 1]
diff_sign = min(pos1_phi) < 0 < max(pos1_phi)
if diff_sign & (np.abs(np.min(pos1_phi)) > np.pi / 2.):
pos1_phi[pos1_phi < 0] = np.pi + np.pi + pos1_phi[pos1_phi < 0]
mean_phi = np.average(pos1_phi, weights=q1)
def build_r_map(input_file: str, output_file: str, threshold: float):
"""
This function extracts the true radial position
and the variance of the SiPM positions
in phi and z for each gamma interaction.
"""
DataSiPM = db.DataSiPMsim_only('petalo', 0) # full body PET
DataSiPM_idx = DataSiPM.set_index('SensorID')
try:
sns_response = pd.read_hdf(input_file, 'MC/sns_response')
except ValueError:
print(f'File {input_file} not found')
exit()
except OSError:
print(f'File {input_file} not found')
exit()
except KeyError:
print(f'No object named MC/sns_response in file {input_file}')
exit()
print(f'Analyzing file {input_file}')
sel_df = rf.find_SiPMs_over_threshold(sns_response, threshold)
particles = pd.read_hdf(input_file, 'MC/particles')
hits = pd.read_hdf(input_file, 'MC/hits')
events = particles.event_id.unique()
true_r1, true_r2 = [], []
var_phi1, var_phi2 = [], []
var_z1, var_z2 = [], []
touched_sipms1, touched_sipms2 = [], []
for evt in events:
### Select photoelectric events only
evt_parts = particles[particles.event_id == evt]
evt_hits = hits[hits.event_id == evt]
select, true_pos = mcf.select_photoelectric(evt_parts, evt_hits)
if not select: continue
sns_resp = sel_df[sel_df.event_id == evt]
if len(sns_resp) == 0: continue
_, _, pos1, pos2, q1, q2 = rf.assign_sipms_to_gammas(
sns_resp, true_pos, DataSiPM_idx)
if len(pos1) > 0:
pos_phi = rf.from_cartesian_to_cyl(np.array(pos1))[:, 1]
_, var_phi = rf.phi_mean_var(pos_phi, q1)
pos_z = np.array(pos1)[:, 2]
mean_z = np.average(pos_z, weights=q1)
var_z = np.average((pos_z - mean_z)**2, weights=q1)
r = np.sqrt(true_pos[0][0]**2 + true_pos[0][1]**2)
var_phi1.append(var_phi)
var_z1.append(var_z)
touched_sipms1.append(len(pos1))
true_r1.append(r)
else:
var_phi1.append(1.e9)
var_z1.append(1.e9)
touched_sipms1.append(1.e9)
true_r1.append(1.e9)
if len(pos2) > 0:
pos_phi = rf.from_cartesian_to_cyl(np.array(pos2))[:, 1]
_, var_phi = rf.phi_mean_var(pos_phi, q2)
pos_z = np.array(pos2)[:, 2]
mean_z = np.average(pos_z, weights=q2)
var_z = np.average((pos_z - mean_z)**2, weights=q2)
r = np.sqrt(true_pos[1][0]**2 + true_pos[1][1]**2)
var_phi2.append(var_phi)
var_z2.append(var_z)
touched_sipms2.append(len(pos2))
true_r2.append(r)
else:
var_phi2.append(1.e9)
var_z2.append(1.e9)
touched_sipms2.append(1.e9)
true_r2.append(1.e9)
a_true_r1 = np.array(true_r1)
a_true_r2 = np.array(true_r2)
a_var_phi1 = np.array(var_phi1)
a_var_phi2 = np.array(var_phi2)
a_var_z1 = np.array(var_z1)
a_var_z2 = np.array(var_z2)
a_touched_sipms1 = np.array(touched_sipms1)
a_touched_sipms2 = np.array(touched_sipms2)
np.savez(output_file,
a_true_r1=a_true_r1,
a_true_r2=a_true_r2,
a_var_phi1=a_var_phi1,
a_var_phi2=a_var_phi2,
a_var_z1=a_var_z1,
a_var_z2=a_var_z2,
a_touched_sipms1=a_touched_sipms1,
a_touched_sipms2=a_touched_sipms2)