def analyse_noise_diode(input_file,output_dir='.',antenna='sd',targets='all',freq_chans=None,rfi_mask=None, nd_models=None):
# Get data from h5 file and use 'select' to obtain a useable subset of it.
data = read_and_select_file(input_file, bline=antenna, channels=freq_chans, rfi_mask=rfi_mask, nd_models=nd_models)
pdf = PdfPages(os.path.join(output_dir,os.path.splitext(os.path.basename(input_file))[0] +'_SystemTemp_'+data.antenna.name+'.pdf'))
#Convert the data to kelvin using the noise diode firings
data.convert_power_to_temperature()
#average each required scan in time sensibly and plot the data for the before and after scans
average_specs=[]
plottitles=[]
compscan_labels=np.array([compscan.label for compscan in data.compscans])
ba_compscans = np.concatenate((np.where(compscan_labels == 'track_before')[0], np.where(compscan_labels == 'track_after')[0],))
for num in ba_compscans:
compscan = data.compscans[num]
compscan_data = np.empty((0,len(data.channel_select),4))
if (targets == 'all') or (compscan.target.name in targets):
for scan in compscan.scans:
scan_data = scan.data[np.where(scan.flags['nd_on']==False)]
compscan_data = np.append(compscan_data,scan_data,axis=0)
average_spec, sigma_spec = robust_mu_sigma(compscan_data, axis=0)
average_specs.append(average_spec[:,:2])
plottitles.append(compscan.target.name + ' ' + compscan_labels[num])
systemp = present_results(pdf, average_specs, data.freqs, plottitles, data.antenna.name, data.bandwidths[0])
#Plot the (before - after) spectrum
#Assume before is first and after second
if len(average_specs)==2:
difftemp = present_difference_results(pdf, average_specs, data.freqs, data.antenna.name, data.bandwidths[0])
else:
print("No before and after tracks. Not plotting difference spectrum.")
#Plot the scan track strong scans
average_specs=[]
plottitles=[]
strong_compscans = np.where(compscan_labels == 'track_strong')[0]
for num in strong_compscans:
compscan = data.compscans[num]
compscan_data = np.empty((0,len(data.channel_select),4))
if (targets == 'all') or (compscan.target.name in targets):
for scan in compscan.scans:
scan_data = scan.data[np.where(scan.flags['nd_on']==False)]
compscan_data = np.append(compscan_data,scan_data,axis=0)
average_spec, sigma_spec = robust_mu_sigma(compscan_data, axis=0)
average_specs.append(average_spec[:,:2])
plottitles.append(compscan.target.name + ' ' + compscan_labels[num])
systemp = present_results(pdf, average_specs, data.freqs, plottitles, data.antenna.name, data.bandwidths[0])
#Get the system temperature in each scan and plot it
alltempshh,alltempsvv,alltimes=[],[],[]
zerotime = data.scans[0].timestamps[0]
for scan in data.scans:
if (targets == 'all') or (scan.compscan.target.name in targets):
scan_data = scan.data[np.where(scan.flags['nd_on']==False)]
average_spec, sigma_spec = robust_mu_sigma(scan_data, axis=0)
systemp=get_system_temp(average_spec[:,:2])
alltempshh.append(systemp[0])
alltempsvv.append(systemp[1])
alltimes.append((np.mean(scan.timestamps[np.where(scan.flags['nd_on']==False)])-zerotime)/(60*60))
plot_temps_time(pdf,alltimes,alltempshh,alltempsvv,data.antenna.name)
pdf.close()
def analyse_noise_diode(input_file,
output_dir='.',
antenna='sd',
targets='all',
freq_chans=None):
# Get data from h5 file and use 'select' to obtain a useable subset of it.
data, compscan_labels = read_and_select_file(input_file,
bline=antenna,
channels=freq_chans)
pdf = PdfPages(
os.path.join(
output_dir,
os.path.splitext(os.path.basename(input_file))[0] +
'_SystemTemp_' + data.antenna.name + '.pdf'))
# loop through compscans in file and get noise diode firings
#nd_data = extract_cal_dataset(data)
#Convert the data to kelvin using the noise diode firings
data.convert_power_to_temperature()
#average each required scan in time sensibly and plot the data
for num, compscan in enumerate(data.compscans):
compscan_data = np.empty((0, len(data.channel_select), 4))
if (targets == 'all') or (compscan.target.name in targets):
for scan in compscan.scans:
scan_data = scan.data[np.where(scan.flags['nd_on'] == False)]
compscan_data = np.append(compscan_data, scan_data, axis=0)
average_spec, sigma_spec = robust_mu_sigma(compscan_data, axis=0)
plottitle = compscan.target.name + ' ' + compscan_labels[num]
systemp = present_results(pdf, average_spec[:, :2],
data.freqs * 1.e6, plottitle,
data.antenna.name,
data.bandwidths[0] * 1e6)
#Get the system temperature in each scan and plot it
alltempshh, alltempsvv, alltimes = [], [], []
zerotime = data.scans[0].timestamps[0]
for scan in data.scans:
if (targets == 'all') or (scan.compscan.target.name in targets):
scan_data = scan.data[np.where(scan.flags['nd_on'] == False)]
average_spec, sigma_spec = robust_mu_sigma(scan_data, axis=0)
systemp = get_system_temp(average_spec[:, :2])
alltempshh.append(systemp[0])
alltempsvv.append(systemp[1])
alltimes.append((np.mean(scan.timestamps[np.where(
scan.flags['nd_on'] == False)]) - zerotime) / (60 * 60))
plot_temps_time(pdf, alltimes, alltempshh, alltempsvv, data.antenna.name)
pdf.close()
def present_results(pdf, temperatures, freq, targnames, antenna, channelwidth):
#Set up the figure
fig = plt.figure(figsize=(8.3,8.3))
flagger=rfilib.sumthreshold_flagger(spike_width_time=1,spike_width_freq=5)
ax = [plt.subplot(211),plt.subplot(212)]
for num,targname in enumerate(targnames):
temperature = temperatures[num]
#Flag the data
flags = flagger.get_flags(np.abs(np.expand_dims(temperature,0))).squeeze()
#Loop over polarisations
temps=[]
for polnum,thispol in enumerate(['HH','VV']):
thisdata = temperature[:,polnum]
thisflags = flags[:,polnum]
systemp, err = robust_mu_sigma(thisdata[np.where(thisflags==False)])
temps.append(systemp)
ax[polnum].set_title('Antenna: ' + antenna + ', ' + thispol + ' pol')
ax[polnum].plot(freq,thisdata,label=targname + ' Tsys: %5.2f'%(systemp))
ax[polnum].set_xlabel('Frequency (MHz)')
ax[polnum].set_ylabel('System Temperature (K)')
ax[polnum].set_xlim(min(freq), max(freq))
ax[0].legend(loc=3)
ax[1].legend(loc=3)
pdf.savefig()
plt.close(fig)
return temps
def present_difference_results(pdf,temperatures, freq, antenna, channelwidth):
#Set up the figure
fig = plt.figure(figsize=(8.3,8.3))
flagger=rfilib.sumthreshold_flagger(spike_width_time=1,spike_width_freq=5)
ax = [plt.subplot(211),plt.subplot(212)]
#ASSUMPTION: before scan is element 0 and after scan is element 1
temperature_before = temperatures[0]
temperature_after = temperatures[1]
#Flag the data
flags_before = flagger.get_flags(np.abs(np.expand_dims(temperature_before,0))).squeeze()
flags_after = flagger.get_flags(np.abs(np.expand_dims(temperature_after,0))).squeeze()
difference = temperature_before-temperature_after
flags = flags_before | flags_after
#Loop over polarisations
temps=[]
for polnum,thispol in enumerate(['HH','VV']):
thisdata = difference[:,polnum]
thisflags = flags[:,polnum]
systemp, err = robust_mu_sigma(thisdata[np.where(thisflags==False)])
temps.append(systemp)
ax[polnum].set_title('Antenna: ' + antenna + ', ' + thispol + ' pol')
ax[polnum].plot(freq[np.where(thisflags==False)],thisdata[np.where(thisflags==False)],label='track_before - track_after, Mean: %5.2f'%(systemp))
ax[polnum].set_xlabel('Frequency (MHz)')
ax[polnum].set_ylabel('System Temperature Difference (K)')
ax[polnum].set_xlim(min(freq), max(freq))
#ax[polnum].set_ylim(min(thisdata[np.where(thisflags==False)]), max(thisdata[np.where(thisflags==False)]))
ax[0].legend(loc=3)
ax[1].legend(loc=3)
pdf.savefig()
plt.close(fig)
return temps
def present_results(pdf, temperature, freq, targname, antenna, channelwidth):
#Set up the figure
fig = plt.figure(figsize=(8.3, 8.3))
#Flag the data
flags = rfilib.detect_spikes_sumthreshold(temperature)
#Loop over polarisations
temps = []
for polnum, thispol in enumerate(['HH', 'VV']):
thisdata = temperature[:, polnum]
thisflags = flags[:, polnum]
systemp, err = robust_mu_sigma(thisdata[np.where(thisflags == False)])
temps.append(systemp)
ax = plt.subplot(211 + polnum)
plt.title(targname + ', Antenna: ' + antenna + ', ' + thispol + ' pol')
ax.text(0.05, 0.8, 'Tsys: %5.2f' % (systemp), transform=ax.transAxes)
ax.plot(freq, thisdata)
plt.xlabel('Frequency (Hz)')
plt.ylabel('System Temperature (K)')
rfilib.plot_RFI_mask(ax,
extra=freq[np.where(thisflags)],
channelwidth=channelwidth)
plt.xlim(freq[-1], freq[0])
pdf.savefig()
plt.close(fig)
return temps
def get_system_temp(temperature):
flagger=rfilib.sumthreshold_flagger(spike_width_time=1,spike_width_freq=5)
flags = flagger.get_flags(np.expand_dims(temperature,0)).squeeze()
temps=[]
for polnum,thispol in enumerate(['HH','VV']):
thisdata= temperature[:,polnum]
thisflags= flags[:,polnum]
systemp,err = robust_mu_sigma(thisdata[np.where(thisflags==False)])
temps.append(systemp)
return temps
def get_system_temp(temperature):
flags = rfilib.detect_spikes_sumthreshold(temperature)
temps = []
for polnum, thispol in enumerate(['HH', 'VV']):
thisdata = temperature[:, polnum]
thisflags = flags[:, polnum]
systemp, err = robust_mu_sigma(thisdata[np.where(thisflags == False)])
temps.append(systemp)
return temps