def energy_transfer_compare_MG_and_He3(df_MG, calibration, Ei, MG_solid_angle,
He3_solid_angle, p0, window):
# Declare parameters
number_bins = int(window.dE_bins.text())
calibration = get_calibration(calibration, Ei)
# Filter Multi-Grid data
df_MG = filter_ce_clusters(window, df_MG)
# Calculate Multi-Grid and He3 energy transfer
dE_MG = get_MG_dE(df_MG, calibration, Ei)
dE_He3 = get_raw_He3_dE(calibration, Ei)
# Histogram dE data
dE_MG_hist, __ = np.histogram(dE_MG, bins=number_bins, range=[-Ei, Ei])
dE_He3_hist, bins = np.histogram(dE_He3, bins=number_bins, range=[-Ei, Ei])
bin_centers = 0.5 * (bins[1:] + bins[:-1])
# Calculate normalization based on accumulated charge and solid angle
MG_charge, He3_charge = get_charge(calibration)
norm_charge_solid_angle = ((He3_solid_angle / MG_solid_angle) *
(He3_charge / MG_charge))
dE_MG_hist_normalized = dE_MG_hist * max(dE_He3_hist) / max(
dE_MG_hist) #* norm_charge_solid_angle
# Calculate elastic peak area ratio and FHWM
MG_peak_area, MG_FWHM, p0 = get_peak_area_and_FWHM(bin_centers,
dE_MG_hist_normalized,
calibration, p0, 'MG',
window)
He3_peak_area, He3_FWHM, p0 = get_peak_area_and_FWHM(
bin_centers, dE_He3_hist, calibration, p0, 'He3', window)
peak_ratio = MG_peak_area / He3_peak_area
# Plot data
fig = plt.figure()
plt.plot(bin_centers,
dE_MG_hist_normalized,
'-',
color='red',
label='Multi-Grid',
zorder=5)
plt.plot(bin_centers,
dE_He3_hist,
'-',
color='blue',
label='$^3$He-tubes',
zorder=5)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized Counts')
plt.yscale('log')
plt.title('Energy transfer%s\nData set: %s' %
(get_detector(window), calibration))
plt.legend(loc=1)
# Export histograms to text-files
export_dE_histograms_to_text(calibration, bin_centers, dE_MG_hist,
dE_He3_hist)
return [fig, peak_ratio, MG_FWHM, He3_FWHM, p0]
def energy_transfer_histogram(df, calibration, Ei, window):
# Declare parameters
number_bins = int(window.dE_bins.text())
calibration = get_calibration(calibration, Ei)
# Filter Multi-Grid data
df = filter_ce_clusters(window, df)
# Calculate Multi-Grid energy transfer
dE = get_MG_dE(df, calibration, Ei)
# Histogram dE data
dE_hist, bins = np.histogram(dE, bins=number_bins, range=[-Ei, Ei])
bin_centers = 0.5 * (bins[1:] + bins[:-1])
# Plot data
fig = plt.figure()
plt.plot(bin_centers, dE_hist, '.-', color='black', zorder=5)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Counts')
plt.yscale('log')
plt.title('Energy transfer\nData set: %s' % calibration)
return fig
def C4H2I2S_compare_all_energies(window):
# Declare folder paths
dir_name = os.path.dirname(__file__)
MG_folder = os.path.join(dir_name, '../../Clusters/MG_old/')
He3_folder = os.path.join(dir_name, '../../Archive/2019_01_10_SEQ_Diiodo/')
# Import file names
MG_files = import_C4H2I2S_MG_files()
He3_files = import_C4H2I2S_He3_files()
# Declare parameters
colors = ['C0', 'C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7']
number_of_bins = 290
Energies = [21, 35, 50.12, 70.13, 99.9, 197.3, 296.1, 492]
calibrations = [
'C4H2I2S_21.0', 'C4H2I2S_35.0', 'C4H2I2S_50.0', 'C4H2I2S_70.0',
'C4H2I2S_100.0', 'C4H2I2S_200.0', 'C4H2I2S_300.0', 'C4H2I2S_500.0'
]
# Iterate through all energies
for MG_file, He3_files_short, Ei, calibration, color in zip(
MG_files, He3_files, Energies, calibrations, colors):
# Import He3 data and histogram dE
He3_dE_hist = np.zeros(number_of_bins)
for file_number, measurement_id in enumerate(He3_files_short):
nxs_file = 'SEQ_' + str(measurement_id) + '_autoreduced.nxs'
nxs_path = He3_folder + nxs_file
nxs = h5py.File(nxs_path, 'r')
he3_bins = (
nxs['mantid_workspace_1']['event_workspace']['axis1'].value)
he3_min = he3_bins[0]
he3_max = he3_bins[-1]
He3_bin_centers = 0.5 * (he3_bins[1:] + he3_bins[:-1])
dE = nxs['mantid_workspace_1']['event_workspace']['tof'].value
He3_dE_hist_small, __ = np.histogram(dE,
bins=number_of_bins,
range=[he3_min, he3_max])
He3_dE_hist += He3_dE_hist_small
# Import MG data, calculate energy transfer and filter
df_MG = pd.read_hdf(MG_folder + MG_file, 'coincident_events')
df_MG = filter_ce_clusters(window, df_MG)
dE_MG = get_MG_dE(df_MG, calibration, Ei)
# Histogram dE data
MG_dE_hist, __ = np.histogram(dE_MG,
bins=number_of_bins,
range=[he3_min, he3_max])
# Normalize MG data
norm = sum(He3_dE_hist) / sum(MG_dE_hist)
MG_dE_hist_normalized = MG_dE_hist * norm
# Plot data
fig = plt.figure()
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.title('C$_4$H$_2$I$_2$S\nE$_i$: %.2f' % Ei)
plt.plot(He3_bin_centers,
MG_dE_hist_normalized,
color=color,
label='Multi-Grid',
zorder=5)
plt.plot(He3_bin_centers,
He3_dE_hist,
linestyle='--',
color='black',
label='$^3$He-tubes',
zorder=5)
plt.legend(loc=1)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized counts')
plt.yscale('log')
file_name = 'C4H2I2S_%d_meV.pdf' % int(Ei)
fig.savefig(
os.path.join(dir_name, '../../Results/C4H2I2S/%s' % file_name))
plt.close()
def analyze_lineshape(window):
def Gaussian(x, a, x0, sigma):
return a * np.exp(-(x - x0)**2 / (2 * sigma**2))
def meV_to_A(energy):
return np.sqrt(81.81 / energy)
def fit_data(bin_centers, dE_hist, p0):
center_idx = len(bin_centers) // 2
zero_idx = find_nearest(bin_centers[center_idx - 10:center_idx + 10],
0)
zero_idx += center_idx - 10
fit_bins = np.arange(zero_idx - 20, zero_idx + 20 + 1, 1)
elastic_peak_max = max(dE_hist[fit_bins])
popt, __ = scipy.optimize.curve_fit(Gaussian,
bin_centers[fit_bins],
dE_hist[fit_bins],
p0=p0)
#plt.plot(bin_centers, Gaussian(bin_centers, popt[0], popt[1], popt[2]))
sigma = abs(popt[2])
x0 = popt[1]
return sigma, x0, elastic_peak_max, popt
# Declare all input-paths
dir_name = os.path.dirname(__file__)
HR_folder = os.path.join(dir_name, '../../Clusters/MG/HR/')
HR_files = np.array(
[file for file in os.listdir(HR_folder) if file[-3:] == '.h5'])
HR_files_sorted = sorted(HR_files, key=lambda element: get_energy(element))
input_paths = append_folder_and_files(HR_folder, HR_files_sorted)
# Declare output folder
output_folder = os.path.join(dir_name, '../../Results/Shoulder/')
# Declare parameters
colors = {
'ILL': 'blue',
'ESS_CLB': 'red',
'ESS_PA': 'green',
'He3': 'black'
}
p0 = None
number_bins = 150
# Declare vectors where results will be stored
data = {
'ILL': {
'Ei': [],
'value': [],
'error': []
},
'ESS_CLB': {
'Ei': [],
'value': [],
'error': []
},
'ESS_PA': {
'Ei': [],
'value': [],
'error': []
},
'He3': {
'Ei': [],
'value': [],
'error': []
}
}
PA_over_CLB_error = np.zeros(len(input_paths[1:-10]))
HR_energies, __ = get_all_energies()
# Iterate through all calibrations
for idx, input_path in enumerate(input_paths[1:-10]):
# Import parameters
E_i = pd.read_hdf(input_path, 'E_i')['E_i'].iloc[0]
calibration = (pd.read_hdf(input_path,
'calibration')['calibration'].iloc[0])
df_MG_temp = pd.read_hdf(input_path, 'coincident_events')
for detector in ['ILL', 'ESS_CLB', 'ESS_PA']:
# Filter Multi-Grid data
df_MG = filter_ce_clusters(window, df_MG_temp)
df_MG_detector = detector_filter(df_MG, detector)
# Get energy transfer data
dE_MG = get_MG_dE(df_MG_detector, calibration, E_i)
dE_He3 = get_raw_He3_dE(calibration, E_i)
# Produce histograms
hist_MG, bins = np.histogram(dE_MG,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
hist_He3, __ = np.histogram(dE_He3,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
bin_centers = 0.5 * (bins[1:] + bins[:-1])
bins_per_dE = number_bins / (2 * (E_i / 5))
# Fit data
MG_values = fit_data(bin_centers, hist_MG, p0)
He3_values = fit_data(bin_centers, hist_He3, p0)
sigma_MG, x0_MG, peak_MG, p0 = MG_values
sigma_He3, x0_He3, peak_He3, p0 = He3_values
# Extract key values
## Get Background
if E_i <= 7:
back_start = 7
back_stop = 13
else:
back_start = 10
back_stop = 16
back_MG_counts = dE_MG[(
(dE_MG < x0_MG + sigma_MG * back_stop) &
(dE_MG > x0_MG + sigma_MG * back_start))].shape[0]
back_MG = back_MG_counts / (sigma_MG * (back_stop - back_start))
back_He3_counts = dE_He3[(
(dE_He3 < x0_He3 + sigma_He3 * back_stop) &
(dE_He3 > x0_He3 + sigma_He3 * back_start))].shape[0]
back_He3 = back_He3_counts / (sigma_He3 * (back_stop - back_start))
## Normalize data
events_MG = dE_MG[((dE_MG < sigma_MG * 3) &
(dE_MG > (-sigma_MG * 3)))].shape[0]
events_He3 = dE_He3[((dE_He3 < sigma_He3 * 3) &
(dE_He3 > (-sigma_He3 * 3)))].shape[0]
peak_counts_MG = events_MG - back_MG * (back_stop -
back_start) * sigma_MG
peak_counts_He3 = events_He3 - back_He3 * (back_stop -
back_start) * sigma_He3
#norm = peak_counts_He3/peak_counts_MG
#plt.axhline(y=back_MG/bins_per_dE, label='MG back',
# color=colors[detector], linestyle=':')
#plt.axhline(y=back_He3/bins_per_dE, label='He3 back', color='black',
# linestyle=':')
#plt.axvline(x=x0_MG + sigma_MG * back_stop, color=colors[detector],
# linestyle='--')
#plt.axvline(x=x0_MG + sigma_MG * back_start, color=colors[detector],
# linestyle='--')
#plt.axvline(x=x0_He3 + sigma_He3 * back_stop, color='black',
# linestyle='--')
#plt.axvline(x=x0_He3 + sigma_He3 * back_start, color='black',
# linestyle='--')
## Get Shoulder
shoulder_start = 3
shoulder_stop = 5
MG_shoulder = dE_MG[(
(dE_MG < x0_MG + sigma_MG * shoulder_stop) &
(dE_MG > x0_MG + sigma_MG * shoulder_start))].shape[0]
MG_shoulder_minus_back = MG_shoulder - back_MG * (
shoulder_stop - shoulder_start) * sigma_MG
He3_shoulder = dE_He3[(
(dE_He3 < x0_He3 + sigma_He3 * shoulder_stop) &
(dE_He3 > x0_He3 + sigma_He3 * shoulder_start))].shape[0]
He3_shoulder_minus_back = He3_shoulder - back_He3 * (
shoulder_stop - shoulder_start) * sigma_He3
MG_FoM = MG_shoulder_minus_back / peak_counts_MG
He3_FoM = He3_shoulder_minus_back / peak_counts_He3
FoM_frac = MG_FoM / He3_FoM
# Get statistical uncertainties
## All measurements and corresponding statistical uncertainites
print('Calibration: %s, detector: %s' % (calibration, detector))
a = MG_shoulder
print('MG shoulder counts: %s' % MG_shoulder)
da = np.sqrt(MG_shoulder)
b = He3_shoulder
print('He3 shoulder counts: %s' % He3_shoulder)
db = np.sqrt(He3_shoulder)
c = back_MG_counts
print('Back MG counts: %s' % back_MG_counts)
dc = np.sqrt(back_MG_counts)
d = back_He3_counts
print('Back He3 counts: %s' % back_He3_counts)
dd = np.sqrt(back_He3_counts)
e = events_MG
print('Peak MG counts: %s' % events_MG)
de = np.sqrt(events_MG)
f = events_He3
print('Peak He3 counts: %s' % events_He3)
df = np.sqrt(events_He3)
## Background removed measurements
g = a - c / 3
dg = np.sqrt(da**2 + (dc / 3)**2)
h = b - d / 3
dh = np.sqrt(db**2 + (dd / 3)**2)
i = e - c
di = np.sqrt(de**2 + (dc)**2)
j = f - d
dj = np.sqrt(df**2 + (dd)**2)
if (detector == 'ESS_CLB') or (detector == 'ESS_PA'):
PA_over_CLB_error[idx] += (dg / g)**2 + (di / i)**2
## Final fraction between MG shoulder and He3 shoulder
MG_FoM_rel_err = np.sqrt((dg / g)**2 + (di / i)**2)
He3_FoM_rel_err = np.sqrt((dh / h)**2 + (dj / j)**2)
FoM_frac_rel_err = np.sqrt((dg / g)**2 + (di / i)**2 +
(dh / h)**2 + (dj / j)**2)
## Absolute uncertainty
MG_FoM_err = MG_FoM_rel_err * MG_FoM
He3_FoM_err = He3_FoM_rel_err * He3_FoM
FoM_frac_err = FoM_frac_rel_err * FoM_frac
# Plot data
#for i in np.arange(1, 6, 1):
# #plt.axvline(x=x0_MG+i*sigma_MG,
# # color='orange', label='%dσ' % i,
# # linestyle=':')
# #plt.axvline(x=x0_MG-i*sigma_MG,
# # color='orange', label=None,
# # linestyle=':')
# plt.axvline(x=x0_He3+i*sigma_He3,
# color='orange', label='%dσ' % i,
# linestyle='-')
# plt.axvline(x=x0_He3-i*sigma_He3,
# color='orange', label=None,
# linestyle='-')
fig = plt.figure()
#plt.axvline(x=x0_He3-3*sigma_He3,
# color='orange', label='-3σ',
# linestyle='-')
#plt.axvline(x=x0_He3)
#plt.axvline(x=x0_He3+3*sigma_He3,
# color='purple', label='3σ',
# linestyle='-')
#plt.axvline(x=x0_He3+5*sigma_He3,
# color='brown', label='5σ',
# linestyle='-')
title = 'Lineshape Investigation\n(%s, %s)' % (calibration,
detector)
plt.title(title)
plt.errorbar(
bin_centers,
hist_He3 / peak_He3,
#np.sqrt(hist_He3)*(1/peak_He3),
fmt='.-',
capsize=5,
zorder=5,
label='$^3$He',
color='black')
plt.errorbar(
bin_centers,
hist_MG / peak_MG,
#np.sqrt(hist_MG)*(1/peak_MG),
fmt='.-',
capsize=5,
zorder=5,
label='MG.SEQ (%s)' % detector,
color=colors[detector])
plt.xlabel('∆E [meV]')
plt.yscale('log')
#plt.ylim([1e-3, 3e-2])
plt.ylabel('Normalized counts')
plt.legend(loc=1)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
#plt.legend(loc=1)
# Save data
output_path = '%s/%s/%s.pdf' % (output_folder, detector,
calibration)
fig.savefig(output_path, bbox_inches='tight')
plt.close()
data[detector]['Ei'].append(E_i)
data[detector]['value'].append(MG_FoM)
data[detector]['error'].append(MG_FoM_err)
# Save text
text_path = '%s/Text/%s/%s.txt' % (output_folder, detector,
calibration)
np.savetxt(text_path,
np.transpose([
bin_centers, hist_He3 / peak_He3, hist_MG / peak_MG
]),
header="bin_centers, He3_normalized, MG_normalized",
delimiter=",")
data['He3']['Ei'].append(E_i)
data['He3']['value'].append(He3_FoM)
print('He3_FoM')
print(He3_FoM)
data['He3']['error'].append(He3_FoM_err)
# Plot summary and save in text files
fig = plt.figure()
plt.title('Lineshape Investigation')
start = 5
labels = [
'Large Grid (⊥ and ‖)', 'Small Grid (⊥ and ‖)', 'Small Grid (⊥)',
'$^3$He-tubes'
]
text_folder = '%s/Text/' % (output_folder)
for i, detector in enumerate(data.keys()):
#average = sum(data[detector]['value'][start:])/len(data[detector]['Ei'][start:])
plt.errorbar(data[detector]['Ei'][start:],
data[detector]['value'][start:],
data[detector]['error'][start:],
fmt='.',
capsize=5,
zorder=2,
linestyle='-',
label=labels[i],
color=colors[detector])
array_to_save = np.array([
data[detector]['Ei'][start:], data[detector]['value'][start:],
data[detector]['error'][start:]
])
np.savetxt(text_folder + detector + '.txt',
np.transpose([
data[detector]['Ei'][start:],
data[detector]['value'][start:],
data[detector]['error'][start:]
]),
header="Ei, FoM, error",
delimiter=",")
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xscale('log')
plt.legend(loc=2)
plt.xlabel('E$_i$ [meV]')
plt.ylabel('FoM (MG/He3)')
plt.tight_layout()
plt.show()
def compare_all_shoulders(window):
def get_statistical_uncertainty(value, a, b, c, d):
da = np.sqrt(a)
db = np.sqrt(b)
dc = np.sqrt(c)
dd = np.sqrt(d)
dA = np.sqrt(da**2 + dc**2)
dB = np.sqrt(db**2 + dd**2)
return np.sqrt((dA / (a - c))**2 + (dB / (c - d))**2) * value
# Declare parameters
number_bins = int(window.dE_bins.text())
sigma_interval = [3, 8]
# Declare all input-paths
dir_name = os.path.dirname(__file__)
HR_folder = os.path.join(dir_name, '../../Clusters/MG_new/HR/')
HR_files = np.array(
[file for file in os.listdir(HR_folder) if file[-3:] == '.h5'])
HR_files_sorted = sorted(HR_files, key=lambda element: get_energy(element))
paths = append_folder_and_files(HR_folder, HR_files_sorted)
# Get output path
temp_folder = os.path.join(dir_name, '../../Results/Shoulder/temp/')
mkdir_p(temp_folder)
# Calculate He3-solid angle
__, He3_solid_angle = get_He3_tubes_area_and_solid_angle()
# Declare vectors where results will be stored
detectors = ['ILL', 'ESS_CLB', 'ESS_PA']
FoMs = {'ILL': [], 'ESS_CLB': [], 'ESS_PA': []}
colors = {'ILL': 'blue', 'ESS_CLB': 'red', 'ESS_PA': 'green'}
energies = []
errors = {'ILL': [], 'ESS_CLB': [], 'ESS_PA': []}
count = 0
# Iterate through all data and calculate FoM
for run in range(2):
for detector in detectors:
for i, path in enumerate(
paths[8:20]): # Only look at 2 meV to 30 meV 8->16
# Import parameters
E_i = pd.read_hdf(path, 'E_i')['E_i'].iloc[0]
calibration = (pd.read_hdf(
path, 'calibration')['calibration'].iloc[0])
df_MG_temp = pd.read_hdf(path, 'coincident_events')
df_MG = filter_ce_clusters(window, df_MG_temp)
# Get MG dE-data
df_MG_detector = detector_filter(df_MG, detector)
dE_MG = get_MG_dE(df_MG_detector, calibration, E_i)
# Get He3 dE-data
dE_He3 = get_raw_He3_dE(calibration, E_i)
# Produce histograms
hist_MG, bins = np.histogram(dE_MG,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
hist_He3, __ = np.histogram(dE_He3,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
bin_centers = 0.5 * (bins[1:] + bins[:-1])
# Fit data
p0 = None
area_MG, FWHM_MG, __, __, __, popt_MG, l_p_idx_MG, r_p_idx_MG, back_MG, MG_back_indices = Gaussian_fit(
bin_centers, hist_MG, p0)
area_He3, FWHM_He3, __, __, __, popt_He3, l_p_idx_He3, r_p_idx_He3, back_He3, He3_back_indices = Gaussian_fit(
bin_centers, hist_He3, p0)
# Normalize data
norm_peak_area = area_He3 / area_MG
normalized_hist_MG = hist_MG * norm_peak_area
normalized_MG_back = back_MG * norm_peak_area
# Compare difference at certain interval
x0_MG = popt_MG[1]
sigma_MG = abs(popt_MG[2])
x0_He3 = popt_He3[1]
sigma_He3 = abs(popt_He3[2])
left_idx_MG = find_nearest(
bin_centers, x0_MG + sigma_interval[0] * sigma_MG)
right_idx_MG = find_nearest(
bin_centers, x0_MG + sigma_interval[1] * sigma_MG)
indices_MG = np.arange(left_idx_MG, right_idx_MG + 1, 1)
left_idx_He3 = find_nearest(
bin_centers, x0_He3 + sigma_interval[0] * sigma_He3)
right_idx_He3 = find_nearest(
bin_centers, x0_He3 + sigma_interval[1] * sigma_He3)
indices_He3 = np.arange(left_idx_He3, right_idx_He3 + 1, 1)
area_MG = sum(normalized_hist_MG[indices_MG]
) - len(indices_MG) * normalized_MG_back
area_He3 = sum(
hist_He3[indices_He3]) - len(indices_He3) * back_He3
FoM = area_MG / area_He3
# Find uncertainty
error = get_statistical_uncertainty(
FoM, sum(hist_He3[indices_He3]), sum(hist_MG[indices_MG]),
sum(hist_MG[MG_back_indices]),
sum(hist_He3[He3_back_indices]))
if run == 0:
FoMs[detector].append(FoM)
errors[detector].append(error)
if detector == 'ILL':
energies.append(E_i)
else:
# Plot data
fig = plt.figure()
#main_title = 'Peak shape comparison $^3$He-tubes and Multi-Grid'
#main_title += '\nInterval: ' + str(sigma_interval) + '$\sigma$'
#fig.suptitle(main_title, x=0.5, y=1.02)
fig.set_figheight(6)
fig.set_figwidth(20)
plt.subplot(1, 3, 1)
plt.plot(bin_centers,
normalized_hist_MG,
color='red',
label='Multi-Grid',
zorder=5)
plt.plot(bin_centers[MG_back_indices],
normalized_hist_MG[MG_back_indices],
color='black',
label=None,
zorder=10)
plt.fill_between(bin_centers[indices_MG],
normalized_hist_MG[indices_MG],
np.ones(len(indices_MG)) *
normalized_MG_back,
facecolor='orange',
label='Shoulder area',
alpha=1,
zorder=2)
plt.title(detector + ': ' + calibration)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized counts')
plt.yscale('log')
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.legend(loc=1)
plt.subplot(1, 3, 2)
plt.plot(bin_centers[He3_back_indices],
hist_He3[He3_back_indices],
color='black',
label=None,
zorder=6)
plt.plot(bin_centers,
hist_He3,
color='blue',
label='$^3$He-tubes',
zorder=5)
plt.fill_between(bin_centers[indices_He3],
hist_He3[indices_He3],
np.ones(len(indices_He3)) * back_He3,
facecolor='purple',
label='Shoulder area',
alpha=1,
zorder=2)
plt.title('$^3$He-tubes: ' + calibration)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized counts')
plt.yscale('log')
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.legend(loc=1)
plt.subplot(1, 3, 3)
for detector_temp in detectors:
plt.errorbar(energies,
FoMs[detector_temp],
errors[detector_temp],
ecolor=colors[detector_temp],
fmt='.',
capsize=5,
color=colors[detector_temp],
label=detector_temp,
zorder=5,
linestyle='-')
# plt.plot(energies, FoMs[detector_temp], '.-',
# color=colors[detector_temp],
# label=detector_temp, zorder=5)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xlabel('$E_i$ [meV]')
plt.ylabel('Figure-of-Merit')
plt.yscale('log')
plt.title('Figure-of-Merit')
if ((detector == 'ESS_PA') and (i == 4)):
pass
else:
plt.axvline(x=energies[i],
color=colors[detector],
zorder=5)
plt.legend(loc=2)
# Save fig
count += 1
output_path = temp_folder + str(count) + '.png'
plt.tight_layout()
fig.savefig(output_path, bbox_inches='tight')
output_path_pdf = os.path.join(
dir_name, '../../Results/pdf_files/%s.pdf' % count)
fig.savefig(output_path_pdf, bbox_inches='tight')
plt.close()
images = []
files = os.listdir(temp_folder)
files = [
file[:-4] for file in files
if file[-9:] != '.DS_Store' and file != '.gitignore'
]
output_path = os.path.join(
dir_name, '../../Results/Shoulder/FOM_sweep_shoulder.gif')
for filename in sorted(files, key=int):
images.append(imageio.imread(temp_folder + filename + '.png'))
imageio.mimsave(output_path, images) #format='GIF', duration=0.33)
shutil.rmtree(temp_folder, ignore_errors=True)
def compare_all_shoulders_5x5(window):
# Declare parameters
number_bins = 150
sigma_interval = [3, 8]
detectors = ['ILL', 'ESS_CLB', 'ESS_PA']
FoMs = {'ILL': [], 'ESS_CLB': [], 'ESS_PA': []}
colors = {'ILL': 'blue', 'ESS_CLB': 'red', 'ESS_PA': 'green'}
energies = []
errors = {'ILL': [], 'ESS_CLB': [], 'ESS_PA': []}
count = 0
# Get input-file paths
dir_name = os.path.dirname(__file__)
folder = os.path.join(dir_name, '../../Clusters/MG_V_5x5_HR/')
files = np.array(
[file for file in os.listdir(folder) if file[-3:] == '.h5'])
files_sorted = sorted(
files,
key=lambda element: float(element[element.find('V_5x5_HR_') + len(
'V_5x5_HR_'):element.find('_meV.h5')]))
paths = np.core.defchararray.add(np.array(len(files_sorted) * [folder]),
files_sorted)
# Get output path
temp_folder = os.path.join(dir_name, '../../Results/Shoulder/temp/')
mkdir_p(temp_folder)
for run in range(2):
for detector in detectors:
for i, path in enumerate(paths):
# Import parameters
E_i = pd.read_hdf(path, 'E_i')['E_i'].iloc[0]
calibration = pd.read_hdf(path,
'calibration')['calibration'].iloc[0]
calibration = get_calibration(calibration, E_i)
df_MG_temp = pd.read_hdf(path, 'coincident_events')
df_MG = filter_ce_clusters(window, df_MG_temp)
# Get MG dE-data
df_MG_detector = detector_filter(df_MG, detector)
dE_MG = get_MG_dE(df_MG_detector, calibration, E_i)
# Get He3 dE-data
dE_He3 = get_raw_He3_dE(calibration, E_i)
# Produce histograms
hist_MG, bins = np.histogram(dE_MG,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
hist_He3, __ = np.histogram(dE_He3,
bins=number_bins,
range=[-E_i / 5, E_i / 5])
bin_centers = 0.5 * (bins[1:] + bins[:-1])
# Normalize data
p0 = None
area_MG, FWHM_MG, __, __, __, popt_MG, l_p_idx_MG, r_p_idx_MG, back_MG = Gaussian_fit(
bin_centers, hist_MG, p0)
area_He3, FWHM_He3, __, __, __, popt_He3, l_p_idx_He3, r_p_idx_He3, back_He3 = Gaussian_fit(
bin_centers, hist_He3, p0)
norm_area = area_He3 / area_MG
normalized_hist_MG = hist_MG * norm_area
# THIS IS JUST TEMPORARY REMOVE AFTER USE
fig = plt.figure()
plt.plot(bin_centers,
normalized_hist_MG,
'-',
color='red',
label='Multi-Grid',
zorder=5)
plt.plot(bin_centers,
hist_He3,
'-',
color='blue',
label='$^3$He-tubes',
zorder=5)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized Counts')
plt.yscale('log')
plt.title(
'Energy transfer - All detectors (5x5 V-sample)\nData set: %s'
% calibration)
plt.legend(loc=1)
output_path_pdf = os.path.join(
dir_name, '../../Results/pdf_files/%s.pdf' % count)
fig.savefig(output_path_pdf, bbox_inches='tight')
plt.close()
# Compare difference at certain interval
x0_MG = popt_MG[1]
sigma_MG = abs(popt_MG[2])
x0_He3 = popt_He3[1]
sigma_He3 = abs(popt_He3[2])
left_idx_MG = find_nearest(
bin_centers, x0_MG + sigma_interval[0] * sigma_MG)
right_idx_MG = find_nearest(
bin_centers, x0_MG + sigma_interval[1] * sigma_MG)
indices_MG = np.arange(left_idx_MG, right_idx_MG + 1, 1)
left_idx_He3 = find_nearest(
bin_centers, x0_He3 + sigma_interval[0] * sigma_He3)
right_idx_He3 = find_nearest(
bin_centers, x0_He3 + sigma_interval[1] * sigma_He3)
indices_He3 = np.arange(left_idx_He3, right_idx_He3 + 1, 1)
area_MG = sum(
normalized_hist_MG[indices_MG]) - len(indices_MG) * back_MG
area_He3 = sum(
hist_He3[indices_He3]) - len(indices_He3) * back_He3
area_diff = sum(normalized_hist_MG[indices_MG] -
hist_He3[indices_MG])
He3_peak_area = sum(
hist_He3[l_p_idx_He3:r_p_idx_He3] -
np.ones(len(bin_centers[l_p_idx_He3:r_p_idx_He3])) *
back_He3)
FoM = area_diff / He3_peak_area
# Find uncertainty
a = sum(hist_MG[indices_MG])
b = sum(hist_He3[indices_MG])
c = sum(hist_He3[l_p_idx_He3:r_p_idx_He3])
d = sum(
np.ones(len(bin_centers[l_p_idx_He3:r_p_idx_He3])) *
back_He3)
da = np.sqrt(a)
db = np.sqrt(b)
dc = np.sqrt(c)
dd = np.sqrt(d)
uncertainty = np.sqrt((np.sqrt(da**2 + db**2) / (a - b))**2 +
(np.sqrt(dc**2 + dd**2) / (c - d))**2)
error = uncertainty * FoM
if run == 0:
FoMs[detector].append(FoM)
errors[detector].append(error)
if detector == 'ILL':
energies.append(E_i)
else:
# Plot data
fig = plt.figure()
#main_title = 'Peak shape comparison $^3$He-tubes and Multi-Grid'
#main_title += '\nInterval: ' + str(sigma_interval) + '$\sigma$'
#fig.suptitle(main_title, x=0.5, y=1.02)
fig.set_figheight(6)
fig.set_figwidth(14)
plt.subplot(1, 2, 1)
plt.plot(bin_centers,
normalized_hist_MG,
color='red',
label='Multi-Grid',
zorder=5)
plt.plot(bin_centers,
hist_He3,
color='blue',
label='$^3$He-tubes',
zorder=5)
plt.fill_between(bin_centers[indices_MG],
normalized_hist_MG[indices_MG],
hist_He3[indices_MG],
facecolor='orange',
label='Shoulder area',
alpha=1,
zorder=2)
plt.fill_between(
bin_centers[l_p_idx_He3:r_p_idx_He3],
hist_He3[l_p_idx_He3:r_p_idx_He3],
np.ones(len(bin_centers[l_p_idx_He3:r_p_idx_He3])) *
back_He3,
facecolor='purple',
label='$^3$He-peak area',
alpha=1,
zorder=2)
plt.title(detector + ': ' + calibration)
plt.xlabel('$E_i$ - $E_f$ [meV]')
plt.ylabel('Normalized counts')
plt.yscale('log')
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.legend(loc=1)
plt.subplot(1, 2, 2)
for detector_temp in detectors:
plt.errorbar(energies,
FoMs[detector_temp],
errors[detector_temp],
ecolor=colors[detector_temp],
fmt='.',
capsize=5,
color=colors[detector_temp],
label=detector_temp,
zorder=5,
linestyle='-')
# plt.plot(energies, FoMs[detector_temp], '.-',
# color=colors[detector_temp],
# label=detector_temp, zorder=5)
plt.grid(True, which='major', linestyle='--', zorder=0)
plt.grid(True, which='minor', linestyle='--', zorder=0)
plt.xlabel('$E_i$ [meV]')
plt.ylabel('Figure-of-Merit')
plt.title('Figure-of-Merit')
if ((detector == 'ESS_PA') and (i == 4)):
pass
else:
plt.axvline(x=energies[i],
color=colors[detector],
zorder=5)
plt.legend(loc=2)
# Save fig
count += 1
output_path = temp_folder + str(count) + '.png'
plt.tight_layout()
fig.savefig(output_path, bbox_inches='tight')
output_path_pdf = os.path.join(
dir_name, '../../Results/pdf_files/%s.pdf' % count)
#fig.savefig(output_path_pdf, bbox_inches='tight')
images = []
files = os.listdir(temp_folder)
files = [
file[:-4] for file in files
if file[-9:] != '.DS_Store' and file != '.gitignore'
]
output_path = os.path.join(
dir_name, '../../Results/Shoulder/FOM_sweep_shoulder.gif')
for filename in sorted(files, key=int):
images.append(imageio.imread(temp_folder + filename + '.png'))
imageio.mimsave(output_path, images) #format='GIF', duration=0.33)
shutil.rmtree(temp_folder, ignore_errors=True)