def TISC_sim(SNR,
threshold,
b_input_delay,
c_input_delay,
num_bits=3,
noise_sigma=32.0,
sample_freq=2600000000.0,
TISC_sample_length=16,
num_samples=80,
upsample=10,
seed=5522684,
draw_flag=0,
digitization_factor=32.0,
output_dir="output/",
boresight=0,
baseline=0,
debug=False):
# Setup
save_output_flag = 0
#if(save_output_flag):
#outfile = str(output_dir+"/test.root")
trigger_flag = 0
#num_bits = 3 # Number of bits available to the digitizer
#num_samples = num_samples*upsample
filter_flag = False
#sample_frequency = 2800000000.0
def_max_sum = 0.0
def_as_max_sum = 0.0
ghi_max_sum = 0.0
ghi_as_max_sum = 0.0
timestep = 1.0 / sample_freq
# Phi sectors have alternating baselines
if (boresight == 0):
abc_impulse_amp = 1.000
def_impulse_amp = 0.835
elif (boresight == 1):
abc_impulse_amp = 0.962
def_impulse_amp = 0.885
# Fill numpy arrays with zeros
a_input_noise = np.zeros(num_samples)
b_input_noise = np.zeros(num_samples)
c_input_noise = np.zeros(num_samples)
d_input_noise = np.zeros(num_samples)
e_input_noise = np.zeros(num_samples)
f_input_noise = np.zeros(num_samples)
time = np.zeros(num_samples)
upsampled_time = np.zeros(num_samples)
a_input_signal = np.zeros(num_samples)
b_input_signal = np.zeros(num_samples)
c_input_signal = np.zeros(num_samples)
d_input_signal = np.zeros(num_samples)
e_input_signal = np.zeros(num_samples)
f_input_signal = np.zeros(num_samples)
a_input_signal_noise = np.zeros(num_samples)
b_input_signal_noise = np.zeros(num_samples)
c_input_signal_noise = np.zeros(num_samples)
d_input_signal_noise = np.zeros(num_samples)
e_input_signal_noise = np.zeros(num_samples)
f_input_signal_noise = np.zeros(num_samples)
a_dig_waveform = np.zeros(num_samples)
b_dig_waveform = np.zeros(num_samples)
c_dig_waveform = np.zeros(num_samples)
d_dig_waveform = np.zeros(num_samples)
e_dig_waveform = np.zeros(num_samples)
f_dig_waveform = np.zeros(num_samples)
empty_list = np.zeros(num_samples)
###################################
# Generate Thermal Noise
a_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 0)
b_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 1)
c_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 2)
d_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 3)
e_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 4)
f_input_noise = generate_noise(num_samples,
noise_sigma,
filter_flag,
seed=seed + 5)
###################################
#####################################
# Determine RMS of noise and signal amplitude
#noise_rms = np.sqrt(np.mean((a_input_noise-noise_mean)**2,))
signal_amp = SNR * 2 * noise_sigma
#####################################
a_input_noise = butter_bandpass_filter(a_input_noise)
b_input_noise = butter_bandpass_filter(b_input_noise)
c_input_noise = butter_bandpass_filter(c_input_noise)
d_input_noise = butter_bandpass_filter(d_input_noise)
e_input_noise = butter_bandpass_filter(e_input_noise)
f_input_noise = butter_bandpass_filter(f_input_noise)
#####################################
# Generate impulse
if (SNR != 0):
# Generate Signal and Amplify
a_input_signal = impulse_gen(num_samples,
upsample,
draw_flag=draw_flag,
output_dir=output_dir)
difference = np.amax(a_input_signal) - np.amin(
a_input_signal) # Get peak to peak voltage
a_input_signal *= (1 / difference) # Normalize input
a_input_signal *= signal_amp # Amplify
#b_input_signal = np.concatenate([a_input_signal[:num_samples+b_input_delay],empty_list[:(-1)*b_input_delay]])
#c_input_signal = np.concatenate([a_input_signal[:num_samples+c_input_delay],empty_list[:(-1)*c_input_delay]])
b_input_signal = np.concatenate([
a_input_signal[-b_input_delay:], empty_list[:(-1) * b_input_delay]
])
c_input_signal = np.concatenate([
a_input_signal[-c_input_delay:], empty_list[:(-1) * c_input_delay]
])
a_input_signal = a_input_signal * abc_impulse_amp
b_input_signal = b_input_signal * abc_impulse_amp
c_input_signal = c_input_signal * abc_impulse_amp
d_input_signal = a_input_signal * def_impulse_amp
e_input_signal = b_input_signal * def_impulse_amp
f_input_signal = c_input_signal * def_impulse_amp
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_input_signal,time,b_input_signal,time,c_input_signal)
plt.plot(time,d_input_signal,time,e_input_signal,time,f_input_signal)
plt.plot(time,g_input_signal,time,h_input_signal,time,i_input_signal)
plt.title("impulse")
plt.show()
"""
# Add the signal to the noise
a_input_signal_noise = np.add(a_input_noise, a_input_signal)
b_input_signal_noise = np.add(b_input_noise, b_input_signal)
c_input_signal_noise = np.add(c_input_noise, c_input_signal)
d_input_signal_noise = np.add(d_input_noise, d_input_signal)
e_input_signal_noise = np.add(e_input_noise, e_input_signal)
f_input_signal_noise = np.add(f_input_noise, f_input_signal)
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_input_signal_noise,time,b_input_signal_noise,time,c_input_signal_noise)
plt.plot(time,d_input_signal_noise,time,e_input_signal_noise,time,f_input_signal_noise)
#plt.plot(time,g_input_signal_noise,time,h_input_signal_noise,time,i_input_signal_noise)
plt.title("impulse+noise")
plt.show()
"""
else:
a_input_signal_noise = a_input_noise
b_input_signal_noise = b_input_noise
c_input_signal_noise = c_input_noise
d_input_signal_noise = d_input_noise
e_input_signal_noise = e_input_noise
f_input_signal_noise = f_input_noise
##########################################
#time = np.linspace(0.0,timestep*num_samples,num_samples)
#plt.plot(time,a_input_noise,time,b_input_noise,time,c_input_noise)
#plt.title("noise")
#plt.show()
##########################################
# Digitized the incoming signal and noise (RITC)
a_dig_waveform = digitize(a_input_signal_noise, num_samples, num_bits,
digitization_factor)
b_dig_waveform = digitize(b_input_signal_noise, num_samples, num_bits,
digitization_factor)
c_dig_waveform = digitize(c_input_signal_noise, num_samples, num_bits,
digitization_factor)
d_dig_waveform = digitize(d_input_signal_noise, num_samples, num_bits,
digitization_factor)
e_dig_waveform = digitize(e_input_signal_noise, num_samples, num_bits,
digitization_factor)
f_dig_waveform = digitize(f_input_signal_noise, num_samples, num_bits,
digitization_factor)
##########################################
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_dig_waveform,time,b_dig_waveform,time,c_dig_waveform)
plt.plot(time,d_dig_waveform,time,e_dig_waveform,time,f_dig_waveform)
#plt.plot(time,g_dig_waveform,time,h_dig_waveform,time,i_dig_waveform)
plt.title("Digitized")
plt.show()
"""
##########################################
# Run the signal through the GLITC module to get trigger
trigger_flag, max_total_sum, best_angle, total_sum = sum_correlate(
num_samples,
a_dig_waveform,
b_dig_waveform,
c_dig_waveform,
d_dig_waveform,
e_dig_waveform,
f_dig_waveform,
baseline,
threshold,
TISC_sample_length,
debug=debug)
#print abc_max_sum
#print def_max_sum
#print ghi_max_sum
#########################################
#dummy = raw_input('Press any key to close')
return max_total_sum, best_angle
def TISC_sim(SNR,threshold,
b_input_delay,c_input_delay,num_bits=3,
noise_sigma=32.0,
sample_freq=2600000000.0,TISC_sample_length=16,
num_samples=80,upsample=10,
seed=5522684,draw_flag=0,digitization_factor=32.0,
output_dir="output/",boresight=0,baseline=0,debug=False):
# Setup
save_output_flag = 0
#if(save_output_flag):
#outfile = str(output_dir+"/test.root")
trigger_flag = 0
#num_bits = 3 # Number of bits available to the digitizer
#num_samples = num_samples*upsample
filter_flag = False
#sample_frequency = 2800000000.0
def_max_sum = 0.0
def_as_max_sum = 0.0
ghi_max_sum = 0.0
ghi_as_max_sum = 0.0
timestep = 1.0/sample_freq
# Phi sectors have alternating baselines
if(boresight==0):
abc_impulse_amp = 1.000
def_impulse_amp = 0.835
elif(boresight==1):
abc_impulse_amp = 0.962
def_impulse_amp = 0.885
# Fill numpy arrays with zeros
a_input_noise = np.zeros(num_samples)
b_input_noise = np.zeros(num_samples)
c_input_noise = np.zeros(num_samples)
d_input_noise = np.zeros(num_samples)
e_input_noise = np.zeros(num_samples)
f_input_noise = np.zeros(num_samples)
time = np.zeros(num_samples)
upsampled_time = np.zeros(num_samples)
a_input_signal = np.zeros(num_samples)
b_input_signal = np.zeros(num_samples)
c_input_signal = np.zeros(num_samples)
d_input_signal = np.zeros(num_samples)
e_input_signal = np.zeros(num_samples)
f_input_signal = np.zeros(num_samples)
a_input_signal_noise = np.zeros(num_samples)
b_input_signal_noise = np.zeros(num_samples)
c_input_signal_noise = np.zeros(num_samples)
d_input_signal_noise = np.zeros(num_samples)
e_input_signal_noise = np.zeros(num_samples)
f_input_signal_noise = np.zeros(num_samples)
a_dig_waveform = np.zeros(num_samples)
b_dig_waveform = np.zeros(num_samples)
c_dig_waveform = np.zeros(num_samples)
d_dig_waveform = np.zeros(num_samples)
e_dig_waveform = np.zeros(num_samples)
f_dig_waveform = np.zeros(num_samples)
empty_list = np.zeros(num_samples)
###################################
# Generate Thermal Noise
a_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+0)
b_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+1)
c_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+2)
d_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+3)
e_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+4)
f_input_noise = generate_noise(num_samples,noise_sigma,filter_flag,seed=seed+5)
###################################
#####################################
# Determine RMS of noise and signal amplitude
#noise_rms = np.sqrt(np.mean((a_input_noise-noise_mean)**2,))
signal_amp = SNR*2*noise_sigma
#####################################
a_input_noise = butter_bandpass_filter(a_input_noise)
b_input_noise = butter_bandpass_filter(b_input_noise)
c_input_noise = butter_bandpass_filter(c_input_noise)
d_input_noise = butter_bandpass_filter(d_input_noise)
e_input_noise = butter_bandpass_filter(e_input_noise)
f_input_noise = butter_bandpass_filter(f_input_noise)
#####################################
# Generate impulse
if (SNR != 0):
# Generate Signal and Amplify
a_input_signal = impulse_gen(num_samples,upsample,draw_flag=draw_flag,output_dir=output_dir)
difference=np.amax(a_input_signal)-np.amin(a_input_signal) # Get peak to peak voltage
a_input_signal *= (1/difference) # Normalize input
a_input_signal *= signal_amp # Amplify
#b_input_signal = np.concatenate([a_input_signal[:num_samples+b_input_delay],empty_list[:(-1)*b_input_delay]])
#c_input_signal = np.concatenate([a_input_signal[:num_samples+c_input_delay],empty_list[:(-1)*c_input_delay]])
b_input_signal = np.concatenate([a_input_signal[-b_input_delay:],empty_list[:(-1)*b_input_delay]])
c_input_signal = np.concatenate([a_input_signal[-c_input_delay:],empty_list[:(-1)*c_input_delay]])
a_input_signal =a_input_signal*abc_impulse_amp
b_input_signal =b_input_signal*abc_impulse_amp
c_input_signal =c_input_signal*abc_impulse_amp
d_input_signal =a_input_signal*def_impulse_amp
e_input_signal =b_input_signal*def_impulse_amp
f_input_signal =c_input_signal*def_impulse_amp
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_input_signal,time,b_input_signal,time,c_input_signal)
plt.plot(time,d_input_signal,time,e_input_signal,time,f_input_signal)
plt.plot(time,g_input_signal,time,h_input_signal,time,i_input_signal)
plt.title("impulse")
plt.show()
"""
# Add the signal to the noise
a_input_signal_noise = np.add(a_input_noise, a_input_signal)
b_input_signal_noise = np.add(b_input_noise, b_input_signal)
c_input_signal_noise = np.add(c_input_noise, c_input_signal)
d_input_signal_noise = np.add(d_input_noise, d_input_signal)
e_input_signal_noise = np.add(e_input_noise, e_input_signal)
f_input_signal_noise = np.add(f_input_noise, f_input_signal)
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_input_signal_noise,time,b_input_signal_noise,time,c_input_signal_noise)
plt.plot(time,d_input_signal_noise,time,e_input_signal_noise,time,f_input_signal_noise)
#plt.plot(time,g_input_signal_noise,time,h_input_signal_noise,time,i_input_signal_noise)
plt.title("impulse+noise")
plt.show()
"""
else:
a_input_signal_noise = a_input_noise
b_input_signal_noise = b_input_noise
c_input_signal_noise = c_input_noise
d_input_signal_noise = d_input_noise
e_input_signal_noise = e_input_noise
f_input_signal_noise = f_input_noise
##########################################
#time = np.linspace(0.0,timestep*num_samples,num_samples)
#plt.plot(time,a_input_noise,time,b_input_noise,time,c_input_noise)
#plt.title("noise")
#plt.show()
##########################################
# Digitized the incoming signal and noise (RITC)
a_dig_waveform = digitize(a_input_signal_noise,num_samples,num_bits,digitization_factor)
b_dig_waveform = digitize(b_input_signal_noise,num_samples,num_bits,digitization_factor)
c_dig_waveform = digitize(c_input_signal_noise,num_samples,num_bits,digitization_factor)
d_dig_waveform = digitize(d_input_signal_noise,num_samples,num_bits,digitization_factor)
e_dig_waveform = digitize(e_input_signal_noise,num_samples,num_bits,digitization_factor)
f_dig_waveform = digitize(f_input_signal_noise,num_samples,num_bits,digitization_factor)
##########################################
"""
time = np.linspace(0.0,timestep*num_samples,num_samples)
plt.plot(time,a_dig_waveform,time,b_dig_waveform,time,c_dig_waveform)
plt.plot(time,d_dig_waveform,time,e_dig_waveform,time,f_dig_waveform)
#plt.plot(time,g_dig_waveform,time,h_dig_waveform,time,i_dig_waveform)
plt.title("Digitized")
plt.show()
"""
##########################################
# Run the signal through the GLITC module to get trigger
trigger_flag, max_total_sum,best_angle,total_sum = sum_correlate(num_samples,a_dig_waveform,b_dig_waveform,c_dig_waveform,d_dig_waveform,e_dig_waveform,f_dig_waveform,baseline,threshold,TISC_sample_length,debug=debug)
#print abc_max_sum
#print def_max_sum
#print ghi_max_sum
#########################################
#dummy = raw_input('Press any key to close')
return max_total_sum,best_angle
for timestep in range(0,num_runs):
event_passed_flag = 0 # Reset value
max_sum = 0
start = timestep*num_samples_per_window
end = start+num_samples_per_window
# Don't run off the end
if(end>=num_samples):
break
if ((timestep % int((simulation_rate/event_rate)) == 0 or SNR[SNR_counter]==0)):
#print "thermal event"
#print len(a_therm[start:end])
#print len(b_therm[start:end])
#print len(c_therm[start:end])
trigger_flag, max_total_sum,best_delays,total_sum = sum_correlate(num_samples_per_window,a_therm[start:end],b_therm[start:end],c_therm[start:end],d_therm[start:end],e_therm[start:end],f_therm[start:end],abc_baseline,threshold,TISC_sample_length=16)
#d1,def_max_sum,as_def_max_sum,def_correlation_mean,d4,d5,as_def_angle,def_angle = sum_correlate(num_samples_per_window,d_therm[start:end],e_therm[start:end],f_therm[start:end],threshold,def_baseline,TISC_sample_length=16,delay_type_flag=1,average_subtract_flag=1,correlation_mean=def_correlation_mean)
#d1,ghi_max_sum,as_ghi_max_sum,ghi_correlation_mean,d4,d5,as_ghi_angle,ghi_angle = sum_correlate(num_samples_per_window,g_therm[start:end],h_therm[start:end],i_therm[start:end],threshold,ghi_baseline,TISC_sample_length=16,delay_type_flag=1,average_subtract_flag=1,correlation_mean=ghi_correlation_mean)
#print as_abc_max_sum
#print as_def_max_sum
#print as_ghi_max_sum
#print len(a[start:end])
#plt.figure(8)
#plt.plot(time[start:end],a_therm[start:end],time[start:end],d_therm[start:end],time[start:end],g_therm[start:end])
#plt.title("Noise")
#plt.show()
#abc_hist.append(abc_max_sum)
#def_hist.append(def_max_sum)
#ghi_hist.append(ghi_max_sum)
#SNR_hist.append(abc_max_sum+def_max_sum+ghi_max_sum)
# Don't run off the end
if (end >= num_samples):
break
if ((timestep % int((simulation_rate / event_rate)) == 0
or SNR[SNR_counter] == 0)):
#print "thermal event"
#print len(a_therm[start:end])
#print len(b_therm[start:end])
#print len(c_therm[start:end])
trigger_flag, max_total_sum, best_delays, total_sum = sum_correlate(
num_samples_per_window,
a_therm[start:end],
b_therm[start:end],
c_therm[start:end],
d_therm[start:end],
e_therm[start:end],
f_therm[start:end],
abc_baseline,
threshold,
TISC_sample_length=16)
#d1,def_max_sum,as_def_max_sum,def_correlation_mean,d4,d5,as_def_angle,def_angle = sum_correlate(num_samples_per_window,d_therm[start:end],e_therm[start:end],f_therm[start:end],threshold,def_baseline,TISC_sample_length=16,delay_type_flag=1,average_subtract_flag=1,correlation_mean=def_correlation_mean)
#d1,ghi_max_sum,as_ghi_max_sum,ghi_correlation_mean,d4,d5,as_ghi_angle,ghi_angle = sum_correlate(num_samples_per_window,g_therm[start:end],h_therm[start:end],i_therm[start:end],threshold,ghi_baseline,TISC_sample_length=16,delay_type_flag=1,average_subtract_flag=1,correlation_mean=ghi_correlation_mean)
#print as_abc_max_sum
#print as_def_max_sum
#print as_ghi_max_sum
#print len(a[start:end])
#plt.figure(8)
#plt.plot(time[start:end],a_therm[start:end],time[start:end],d_therm[start:end],time[start:end],g_therm[start:end])
#plt.title("Noise")
#plt.show()
