Second Round of Full Day RFI Flagging¶
by Josh Dillon, last updated October 13, 2024
This notebook is synthesizes information from individual delay_filtered_average_zscore notebooks to find low-level RFI and flag it. That notebook takes smooth_calibrated data, redundantly averages it, performs a high-pass delay filter, and then incoherently averages across baselines, creating a per-polarization z-score. This notebook then takes that whole night of z-scores and finds a new set of flags to both add to the smooth_cal files, which are updated in place, and to write down as new UVFlag waterfall-type .h5 files.
Here's a set of links to skip to particular figures and tables:
• Figure 1: Waterfall of Maximum z-Score of Either Polarization Before Round 2 Flagging¶
• Figure 2: Histogram of z-scores¶
• Figure 3: Waterfall of Maximum z-Score of Either Polarization After Round 2 Flagging¶
• Figure 4: Spectra of Time-Averaged z-Scores¶
• Figure 5: Summary of Flags Before and After Round 2 Flagging¶
In [1]:
import time
tstart = time.time()
In [2]:
import os
os.environ['HDF5_USE_FILE_LOCKING'] = 'FALSE'
import h5py
import hdf5plugin # REQUIRED to have the compression plugins available
import numpy as np
import glob
import matplotlib.pyplot as plt
import matplotlib
import copy
import warnings
from pyuvdata import UVFlag, UVCal
from hera_cal import utils
from hera_qm import xrfi
from hera_qm.time_series_metrics import true_stretches
from hera_filters import dspec
from IPython.display import display, HTML
%matplotlib inline
display(HTML("<style>.container { width:100% !important; }</style>"))
_ = np.seterr(all='ignore') # get rid of red warnings
%config InlineBackend.figure_format = 'retina'
In [3]:
# get input data file names
SUM_FILE = os.environ.get("SUM_FILE", None)
# SUM_FILE = '/lustre/aoc/projects/hera/h6c-analysis/IDR2/2459861/zen.2459861.25297.sum.uvh5'
SUM_SUFFIX = os.environ.get("SUM_SUFFIX", 'sum.uvh5')
# get input and output suffixes
SMOOTH_CAL_SUFFIX = os.environ.get("SMOOTH_CAL_SUFFIX", 'sum.smooth.calfits')
ZSCORE_SUFFIX = os.environ.get("ZSCORE_SUFFIX", 'sum.red_avg_zscore.h5')
FLAG_WATERFALL2_SUFFIX = os.environ.get("FLAG_WATERFALL2_SUFFIX", 'sum.flag_waterfall_round_2.h5')
OUT_YAML_SUFFIX = os.environ.get("OUT_YAML_SUFFIX", '_aposteriori_flags.yaml')
OUT_YAML_DIR = os.environ.get("OUT_YAML_DIR", None)
# build globs
sum_glob = '.'.join(SUM_FILE.split('.')[:-3]) + '.*.' + SUM_SUFFIX
cal_files_glob = sum_glob.replace(SUM_SUFFIX, SMOOTH_CAL_SUFFIX)
zscore_glob = sum_glob.replace(SUM_SUFFIX, ZSCORE_SUFFIX)
# build out yaml file
if OUT_YAML_DIR is None:
OUT_YAML_DIR = os.path.dirname(SUM_FILE)
out_yaml_file = os.path.join(OUT_YAML_DIR, SUM_FILE.split('.')[-4] + OUT_YAML_SUFFIX)
# get flagging parameters
Z_THRESH = float(os.environ.get("Z_THRESH", 4))
WS_Z_THRESH = float(os.environ.get("WS_Z_THRESH", 2))
AVG_Z_THRESH = float(os.environ.get("AVG_Z_THRESH", 1))
MAX_FREQ_FLAG_FRAC = float(os.environ.get("MAX_FREQ_FLAG_FRAC", .25))
MAX_TIME_FLAG_FRAC = float(os.environ.get("MAX_TIME_FLAG_FRAC", .1))
AVG_SPECTRUM_FILTER_DELAY = float(os.environ.get("AVG_SPECTRUM_FILTER_DELAY", 250)) # in ns
EIGENVAL_CUTOFF = float(os.environ.get("EIGENVAL_CUTOFF", 1e-12))
TIME_AVG_DELAY_FILT_SNR_THRESH = float(os.environ.get("TIME_AVG_DELAY_FILT_SNR_THRESH", 4.0))
TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE = float(os.environ.get("TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE", 1.5))
for setting in ['Z_THRESH', 'WS_Z_THRESH', 'AVG_Z_THRESH', 'MAX_FREQ_FLAG_FRAC', 'MAX_TIME_FLAG_FRAC', 'AVG_SPECTRUM_FILTER_DELAY',
'EIGENVAL_CUTOFF', 'TIME_AVG_DELAY_FILT_SNR_THRESH', 'TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE']:
print(f'{setting} = {eval(setting)}')
Z_THRESH = 4.0 WS_Z_THRESH = 2.0 AVG_Z_THRESH = 1.0 MAX_FREQ_FLAG_FRAC = 0.25 MAX_TIME_FLAG_FRAC = 0.1 AVG_SPECTRUM_FILTER_DELAY = 250.0 EIGENVAL_CUTOFF = 1e-12 TIME_AVG_DELAY_FILT_SNR_THRESH = 4.0 TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE = 1.5
Load z-scores¶
In [4]:
# load z-scores
zscore_files = sorted(glob.glob(zscore_glob))
print(f'Found {len(zscore_files)} *.{ZSCORE_SUFFIX} files starting with {zscore_files[0]}.')
uvf = UVFlag(zscore_files, use_future_array_shapes=True)
Found 1851 *.sum.red_avg_zscore.h5 files starting with /mnt/sn1/data1/2460753/zen.2460753.25241.sum.red_avg_zscore.h5.
In [5]:
# get calibration solution files
cal_files = sorted(glob.glob(cal_files_glob))
print(f'Found {len(cal_files)} *.{SMOOTH_CAL_SUFFIX} files starting with {cal_files[0]}.')
Found 1851 *.sum.smooth.calfits files starting with /mnt/sn1/data1/2460753/zen.2460753.25241.sum.smooth.calfits.
In [6]:
assert len(zscore_files) == len(cal_files)
In [7]:
# extract z-scores and correct by a single number per polarization to account for biases created by filtering
zscore = {pol: uvf.metric_array[:, :, np.argwhere(uvf.polarization_array == utils.polstr2num(pol, x_orientation=uvf.x_orientation))[0][0]] for pol in ['ee', 'nn']}
zscore = {pol: zscore[pol] - np.nanmedian(zscore[pol]) for pol in zscore}
In [8]:
freqs = uvf.freq_array
times = uvf.time_array
In [9]:
extent = [freqs[0] / 1e6, freqs[-1] / 1e6, times[-1] - int(times[0]), times[0] - int(times[0])]
In [10]:
def plot_max_z_score(zscore, flags=None, vmin=-5, vmax=5):
if flags is None:
flags = np.any(~np.isfinite(list(zscore.values())), axis=0)
plt.figure(figsize=(14,10), dpi=100)
plt.imshow(np.where(flags, np.nan, np.nanmax([zscore['ee'], zscore['nn']], axis=0)), aspect='auto',
cmap='coolwarm', interpolation='none', vmin=vmin, vmax=vmax, extent=extent)
plt.colorbar(location='top', label='Max z-score of either polarization', extend='both', aspect=40, pad=.02)
plt.xlabel('Frequency (MHz)')
plt.ylabel(f'JD - {int(times[0])}')
plt.tight_layout()
Figure 1: Waterfall of Maximum z-Score of Either Polarization Before Round 2 Flagging¶
Shows the worse of the two results from delay_filtered_average_zscore from either polarization. Dips near flagged channels are expected, due to overfitting of noise. Positive-going excursions are problematic and likely evidence of RFI.
In [11]:
plot_max_z_score(zscore)
All-NaN axis encountered
In [12]:
def plot_histogram():
plt.figure(figsize=(14,4), dpi=100)
bins = np.arange(-50, 100, .1)
hist_ee = plt.hist(np.ravel(zscore['ee']), bins=bins, density=True, label='ee-polarized z-scores', alpha=.5)
hist_nn = plt.hist(np.ravel(zscore['nn']), bins=bins, density=True, label='nn-polarized z-scores', alpha=.5)
plt.plot(bins, (2*np.pi)**-.5 * np.exp(-bins**2 / 2), 'k:', label='Gaussian approximate\nnoise-only distribution')
plt.axvline(WS_Z_THRESH, c='r', ls='--', label='Watershed z-score')
plt.axvline(Z_THRESH, c='r', ls='-', label='Threshold z-score')
plt.yscale('log')
all_densities = np.concatenate([hist_ee[0][hist_ee[0] > 0], hist_nn[0][hist_nn[0] > 0]])
plt.ylim(np.min(all_densities) / 2, np.max(all_densities) * 2)
plt.xlim([-50, 100])
plt.legend()
plt.xlabel('z-score')
plt.ylabel('Density')
plt.tight_layout()
Figure 2: Histogram of z-scores¶
Shows a comparison of the histogram of z-scores in this file (one per polarization) to a Gaussian approximation of what one might expect from thermal noise. Without filtering, the actual distribution is a weighted sum of Rayleigh distributions. Filtering further complicates this. To make the z-scores more reliable, a single per-polarization median is subtracted from each waterfall, which allows us to flag low-level outliers with more confidence. Any points beyond the solid red line are flagged. Any points neighboring a flag beyond the dashed red line are also flagged. Finally, flagging is performed for low-level outliers in whole times or channels.
In [13]:
plot_histogram()
Perform flagging¶
In [14]:
def iteratively_flag_on_averaged_zscore(flags, zscore, avg_func=np.nanmean, avg_z_thresh=AVG_Z_THRESH, verbose=True):
'''Flag whole integrations or channels based on average z-score. This is done
iteratively to prevent bad times affecting channel averages or vice versa.'''
flagged_chan_count = 0
flagged_int_count = 0
while True:
zspec = avg_func(np.where(flags, np.nan, zscore), axis=0)
ztseries = avg_func(np.where(flags, np.nan, zscore), axis=1)
if (np.nanmax(zspec) < avg_z_thresh) and (np.nanmax(ztseries) < avg_z_thresh):
break
if np.nanmax(zspec) >= np.nanmax(ztseries):
flagged_chan_count += np.sum((zspec >= np.nanmax(ztseries)) & (zspec >= avg_z_thresh))
flags[:, (zspec >= np.nanmax(ztseries)) & (zspec >= avg_z_thresh)] = True
else:
flagged_int_count += np.sum((ztseries >= np.nanmax(zspec)) & (ztseries >= avg_z_thresh))
flags[(ztseries >= np.nanmax(zspec)) & (ztseries >= avg_z_thresh), :] = True
if verbose:
print(f'\tFlagging an additional {flagged_int_count} integrations and {flagged_chan_count} channels.')
def impose_max_chan_flag_frac(flags, max_flag_frac=MAX_FREQ_FLAG_FRAC, verbose=True):
'''Flag channels already flagged more than max_flag_frac (excluding completely flagged times).'''
unflagged_times = ~np.all(flags, axis=1)
frequently_flagged_chans = np.mean(flags[unflagged_times, :], axis=0) >= max_flag_frac
if verbose:
print(f'\tFlagging {np.sum(frequently_flagged_chans) - np.sum(np.all(flags, axis=0))} channels previously flagged {max_flag_frac:.2%} or more.')
flags[:, frequently_flagged_chans] = True
def impose_max_time_flag_frac(flags, max_flag_frac=MAX_TIME_FLAG_FRAC, verbose=True):
'''Flag times already flagged more than max_flag_frac (excluding completely flagged channels).'''
unflagged_chans = ~np.all(flags, axis=0)
frequently_flagged_times = np.mean(flags[:, unflagged_chans], axis=1) >= max_flag_frac
if verbose:
print(f'\tFlagging {np.sum(frequently_flagged_times) - np.sum(np.all(flags, axis=1))} times previously flagged {max_flag_frac:.2%} or more.')
flags[frequently_flagged_times, :] = True
def time_avg_zscore_dly_filt_SNRs(flags, filter_delay=AVG_SPECTRUM_FILTER_DELAY, eigenval_cutoff=EIGENVAL_CUTOFF):
"""Produces SNRs after time-averaging z-scores and delay filtering, accounting for flagging's effect on the filter."""
# figure out high and low band based on FM gap at 100 MHz
flagged_stretches = true_stretches(np.all(flags, axis=0))
FM_gap = [fs for fs in flagged_stretches if fs.start <= np.argmin(np.abs(freqs - 100e6)) < fs.stop][0]
low_band = slice((0 if flagged_stretches[0].start != 0 else flagged_stretches[0].stop), FM_gap.start)
high_band = slice(FM_gap.stop, (len(freqs) if flagged_stretches[-1].stop != len(freqs) else flagged_stretches[-1].start))
filt_SNR = {}
for pol in zscore:
# calculate timeavg_SNR and filter
noise_prediction = 1.0 / np.sum(~flags, axis=0)**.5
timeavg_SNR = np.nanmean(np.where(flags, np.nan, zscore[pol] / noise_prediction), axis=0)
wgts = np.where(np.isfinite(timeavg_SNR), 1, 0)
model = np.zeros_like(timeavg_SNR)
for band in [low_band, high_band]:
model[band], _, _ = dspec.fourier_filter(freqs[band], np.where(np.isfinite(timeavg_SNR[band]), timeavg_SNR[band], 0),
wgts[band], [0], [AVG_SPECTRUM_FILTER_DELAY / 1e9], mode="dpss_solve",
eigenval_cutoff=[EIGENVAL_CUTOFF], suppression_factors=[EIGENVAL_CUTOFF])
filt_SNR[pol] = timeavg_SNR - model
# correct for impact of filter
correction_factors = np.ones_like(wgts) * np.nan
for band in [low_band, high_band]:
X = dspec.dpss_operator(freqs[band], [0], filter_half_widths=[AVG_SPECTRUM_FILTER_DELAY / 1e9], eigenval_cutoff=[EIGENVAL_CUTOFF])[0]
W = wgts[band]
leverage = np.diag(X @ np.linalg.pinv(np.dot(X.T * W, X)) @ (X.T * W))
correction_factors[band] = np.where(leverage > 0, (1 - leverage)**.5, np.nan) # because the underlying data should be gaussian
filt_SNR[pol] /= correction_factors
return filt_SNR
def iteratively_flag_on_delay_filtered_time_avg_zscore(flags, thresh=TIME_AVG_DELAY_FILT_SNR_THRESH, dynamic_range=TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE,
filter_delay=AVG_SPECTRUM_FILTER_DELAY, eigenval_cutoff=EIGENVAL_CUTOFF):
"""Flag whole channels based on their outlierness after delay-filterd time-averaged zscores.
This is done iteratively since the delay filter can be unduly influenced by large outliers."""
filt_SNR = time_avg_zscore_dly_filt_SNRs(flags, filter_delay=AVG_SPECTRUM_FILTER_DELAY, eigenval_cutoff=EIGENVAL_CUTOFF)
while True:
largest_SNR = np.nanmax(list(filt_SNR.values()))
if largest_SNR < thresh:
break
#
cut = np.max([thresh, largest_SNR / dynamic_range])
for pol in filt_SNR:
flags[:, filt_SNR[pol] > cut] = True
filt_SNR = time_avg_zscore_dly_filt_SNRs(flags, filter_delay=AVG_SPECTRUM_FILTER_DELAY, eigenval_cutoff=EIGENVAL_CUTOFF)
In [15]:
flags = np.any(~np.isfinite(list(zscore.values())), axis=0)
print(f'{np.mean(flags):.3%} of waterfall flagged to start.')
# flag whole integrations or channels using outliers in median
while True:
nflags = np.sum(flags)
for pol in ['ee', 'nn']:
iteratively_flag_on_averaged_zscore(flags, zscore[pol], avg_func=np.nanmedian, avg_z_thresh=AVG_Z_THRESH, verbose=True)
impose_max_chan_flag_frac(flags, max_flag_frac=MAX_FREQ_FLAG_FRAC, verbose=True)
impose_max_time_flag_frac(flags, max_flag_frac=MAX_TIME_FLAG_FRAC, verbose=True)
if np.sum(flags) == nflags:
break
print(f'{np.mean(flags):.3%} of waterfall flagged after flagging whole times and channels with median z > {AVG_Z_THRESH}.')
# flag largest outliers
for pol in ['ee', 'nn']:
flags |= (zscore[pol] > Z_THRESH)
print(f'{np.mean(flags):.3%} of waterfall flagged after flagging z > {Z_THRESH} outliers.')
# watershed flagging
while True:
nflags = np.sum(flags)
for pol in ['ee', 'nn']:
flags |= xrfi._ws_flag_waterfall(zscore[pol], flags, WS_Z_THRESH)
if np.sum(flags) == nflags:
break
print(f'{np.mean(flags):.3%} of waterfall flagged after watershed flagging on z > {WS_Z_THRESH} neighbors of prior flags.')
# flag whole integrations or channels using outliers in mean
while True:
nflags = np.sum(flags)
for pol in ['ee', 'nn']:
iteratively_flag_on_averaged_zscore(flags, zscore[pol], avg_func=np.nanmean, avg_z_thresh=AVG_Z_THRESH, verbose=True)
impose_max_chan_flag_frac(flags, max_flag_frac=MAX_FREQ_FLAG_FRAC, verbose=True)
impose_max_time_flag_frac(flags, max_flag_frac=MAX_TIME_FLAG_FRAC, verbose=True)
if np.sum(flags) == nflags:
break
print(f'{np.mean(flags):.3%} of waterfall flagged after flagging whole times and channels with average z > {AVG_Z_THRESH}.')
# flag channels based on delay filter
iteratively_flag_on_delay_filtered_time_avg_zscore(flags, thresh=TIME_AVG_DELAY_FILT_SNR_THRESH, dynamic_range=TIME_AVG_DELAY_FILT_SNR_DYNAMIC_RANGE,
filter_delay=AVG_SPECTRUM_FILTER_DELAY, eigenval_cutoff=EIGENVAL_CUTOFF)
print(f'{np.mean(flags):.3%} of flagging channels that are {TIME_AVG_DELAY_FILT_SNR_THRESH}σ outliers after delay filtering the time average.')
# watershed flagging again
while True:
nflags = np.sum(flags)
for pol in ['ee', 'nn']:
flags |= xrfi._ws_flag_waterfall(zscore[pol], flags, WS_Z_THRESH)
if np.sum(flags) == nflags:
break
print(f'{np.mean(flags):.3%} of waterfall flagged after another round of watershed flagging on z > {WS_Z_THRESH} neighbors of prior flags.')
22.237% of waterfall flagged to start.
All-NaN slice encountered
Flagging an additional 37 integrations and 10 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more.
Flagging an additional 2 integrations and 6 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more.
Flagging an additional 0 integrations and 0 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more.
Flagging an additional 0 integrations and 0 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more. 23.984% of waterfall flagged after flagging whole times and channels with median z > 1.0. 25.600% of waterfall flagged after flagging z > 4.0 outliers.
27.731% of waterfall flagged after watershed flagging on z > 2.0 neighbors of prior flags. Flagging an additional 0 integrations and 0 channels. Flagging 19 channels previously flagged 25.00% or more. Flagging 546 times previously flagged 10.00% or more.
Mean of empty slice Mean of empty slice
Flagging an additional 0 integrations and 0 channels. Flagging 4 channels previously flagged 25.00% or more. Flagging 1 times previously flagged 10.00% or more. Flagging an additional 0 integrations and 0 channels. Flagging 0 channels previously flagged 25.00% or more.
Flagging 0 times previously flagged 10.00% or more. Flagging an additional 0 integrations and 0 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more. 38.590% of waterfall flagged after flagging whole times and channels with average z > 1.0.
Mean of empty slice Casting complex values to real discards the imaginary part Casting complex values to real discards the imaginary part
41.495% of flagging channels that are 4.0σ outliers after delay filtering the time average.
41.906% of waterfall flagged after another round of watershed flagging on z > 2.0 neighbors of prior flags.
Show results of flagging¶
Figure 3: Waterfall of Maximum z-Score of Either Polarization After Round 2 Flagging¶
The same as Figure 1, but after the flagging performed in this notebook.
In [16]:
plot_max_z_score(zscore, flags=flags)
All-NaN axis encountered
In [17]:
def zscore_spectra(ylim=[-3, 3], flags=flags):
fig, axes = plt.subplots(2, 1, figsize=(14,6), dpi=100, sharex=True, sharey=True, gridspec_kw={'hspace': 0})
for ax, pol in zip(axes, ['ee', 'nn']):
ax.plot(freqs / 1e6, np.nanmean(zscore[pol], axis=0),'r', label=f'{pol}-Polarization Before Round 2 Flagging', lw=.5)
ax.plot(freqs / 1e6, np.nanmean(np.where(flags, np.nan, zscore[pol]), axis=0), label=f'{pol}-Polarization After Round 2 Flagging')
ax.legend(loc='lower right')
ax.set_ylabel('Time-Averged Z-Score\n(Excluding Flags)')
ax.set_ylim(ylim)
axes[1].set_xlabel('Frequency (MHz)')
plt.tight_layout()
Figure 4: Spectra of Time-Averaged z-Scores¶
The average along the time axis of Figures 1 and 3 (though now separated per-polarization). This plot is useful for showing channels with repeated low-level RFI.
In [18]:
zscore_spectra()
Mean of empty slice Mean of empty slice
In [19]:
def summarize_flagging(flags=flags):
plt.figure(figsize=(14,10), dpi=100)
cmap = matplotlib.colors.ListedColormap(((0, 0, 0),) + matplotlib.cm.get_cmap("Set2").colors[0:2])
plt.imshow(np.where(np.any(~np.isfinite(list(zscore.values())), axis=0), 1, np.where(flags, 2, 0)),
aspect='auto', cmap=cmap, interpolation='none', extent=extent)
plt.clim([-.5, 2.5])
cbar = plt.colorbar(location='top', aspect=40, pad=.02)
cbar.set_ticks([0, 1, 2])
cbar.set_ticklabels(['Unflagged', 'Previously Flagged', 'Flagged Here Using Delayed Filtered z-Scores'])
plt.xlabel('Frequency (MHz)')
plt.ylabel(f'JD - {int(times[0])}')
plt.tight_layout()
Figure 5: Summary of Flags Before and After Round 2 Flagging¶
This plot shows which times and frequencies were flagged before and after this notebook. It is directly comparable to Figure 5 of the first round full_day_rfi notebook.
In [20]:
summarize_flagging()
The get_cmap function was deprecated in Matplotlib 3.7 and will be removed in 3.11. Use ``matplotlib.colormaps[name]`` or ``matplotlib.colormaps.get_cmap()`` or ``pyplot.get_cmap()`` instead.
Save results¶
In [21]:
add_to_history = 'by full_day_rfi_round_2 notebook with the following environment:\n' + '=' * 65 + '\n' + os.popen('conda env export').read() + '=' * 65
In [22]:
tind = 0
always_flagged_ants = set()
ever_unflagged_ants = set()
for cal_file in cal_files:
with warnings.catch_warnings():
warnings.simplefilter("ignore")
# update cal_file
uvc = UVCal()
uvc.read(cal_file, use_future_array_shapes=True)
uvc.flag_array |= (flags[tind:tind + len(uvc.time_array), :].T)[None, :, :, None]
uvc.history += 'Modified ' + add_to_history
uvc.write_calfits(cal_file, clobber=True)
# keep track of flagged antennas
for antnum in uvc.ant_array:
for antpol in ['Jee', 'Jnn']:
if np.all(uvc.get_flags(antnum, antpol)):
if (antnum, antpol) not in ever_unflagged_ants:
always_flagged_ants.add((antnum, antpol))
else:
ever_unflagged_ants.add((antnum, antpol))
always_flagged_ants.discard((antnum, antpol))
# Create new flag object
uvf_out = UVFlag(uvc, waterfall=True, mode='flag')
uvf_out.flag_array |= flags[tind:tind + len(uvc.time_array), :, None]
uvf_out.history += 'Produced ' + add_to_history
uvf_out.write(cal_file.replace(SMOOTH_CAL_SUFFIX, FLAG_WATERFALL2_SUFFIX), clobber=True)
# increment time index
tind += len(uvc.time_array)
print(f'Saved {len(cal_files)} *.{FLAG_WATERFALL2_SUFFIX} files starting with {cal_files[0].replace(SMOOTH_CAL_SUFFIX, FLAG_WATERFALL2_SUFFIX)}.')
Saved 1851 *.sum.flag_waterfall_round_2.h5 files starting with /mnt/sn1/data1/2460753/zen.2460753.25241.sum.flag_waterfall_round_2.h5.
In [23]:
# write summary of entirely flagged times/freqs/ants to yaml
all_flagged_times = np.all(flags, axis=1)
all_flagged_freqs = np.all(flags, axis=0)
all_flagged_ants = sorted(always_flagged_ants)
dt = np.median(np.diff(times))
out_yml_str = 'JD_flags: ' + str([[times[flag_stretch][0] - dt / 2, times[flag_stretch][-1] + dt / 2]
for flag_stretch in true_stretches(all_flagged_times)])
df = np.median(np.diff(freqs))
out_yml_str += '\n\nfreq_flags: ' + str([[freqs[flag_stretch][0] - df / 2, freqs[flag_stretch][-1] + df / 2]
for flag_stretch in true_stretches(all_flagged_freqs)])
out_yml_str += '\n\nex_ants: ' + str(all_flagged_ants).replace("'", "").replace('(', '[').replace(')', ']')
print(f'Writing the following to {out_yaml_file}\n' + '-' * (25 + len(out_yaml_file)))
print(out_yml_str)
with open(out_yaml_file, 'w') as outfile:
outfile.writelines(out_yml_str)
Writing the following to /mnt/sn1/data1/2460753/2460753_aposteriori_flags.yaml ------------------------------------------------------------------------------ JD_flags: [[np.float64(2460753.252300107), np.float64(2460753.258675449)], [np.float64(2460753.2590109934), np.float64(2460753.259682082)], [np.float64(2460753.2599057783), np.float64(2460753.260800563)], [np.float64(2460753.260912411), np.float64(2460753.2614716515)], [np.float64(2460753.262030892), np.float64(2460753.263708614)], [np.float64(2460753.2641560063), np.float64(2460753.2646033987)], [np.float64(2460753.264938943), np.float64(2460753.266169272)], [np.float64(2460753.2685180823), np.float64(2460753.2688536267)], [np.float64(2460753.272656462), np.float64(2460753.2727683103)], [np.float64(2460753.2728801584), np.float64(2460753.2729920065)], [np.float64(2460753.27455788), np.float64(2460753.275452665)], [np.float64(2460753.277689627), np.float64(2460753.2781370194)], [np.float64(2460753.279591045), np.float64(2460753.279926589)], [np.float64(2460753.2822753997), np.float64(2460753.2837294247)], [np.float64(2460753.284176817), np.float64(2460753.28518345)], [np.float64(2460753.285742691), np.float64(2460753.285854539)], [np.float64(2460753.2863019314), np.float64(2460753.28697302)], [np.float64(2460753.2875322606), np.float64(2460753.287979653)], [np.float64(2460753.292229881), np.float64(2460753.2925654254)], [np.float64(2460753.294466843), np.float64(2460753.2948023872)], [np.float64(2460753.2951379316), np.float64(2460753.2960327165)], [np.float64(2460753.2979341345), np.float64(2460753.298381527)], [np.float64(2460753.3014014256), np.float64(2460753.30173697)], [np.float64(2460753.30352654), np.float64(2460753.303638388)], [np.float64(2460753.303750236), np.float64(2460753.3040857804)], [np.float64(2460753.3147113505), np.float64(2460753.315270591)], [np.float64(2460753.316277224), np.float64(2460753.316612768)], [np.float64(2460753.3207511483), np.float64(2460753.3210866926)], [np.float64(2460753.3218696294), np.float64(2460753.3220933257)], [np.float64(2460753.323771047), np.float64(2460753.323994743)], [np.float64(2460753.324889528), np.float64(2460753.3252250724)], [np.float64(2460753.329922693), np.float64(2460753.330034541)], [np.float64(2460753.3312648702), np.float64(2460753.3313767184)], [np.float64(2460753.3316004146), np.float64(2460753.332047807)], [np.float64(2460753.332159655), np.float64(2460753.3323833514)], [np.float64(2460753.337416516), np.float64(2460753.3377520605)], [np.float64(2460753.3378639086), np.float64(2460753.3379757567)], [np.float64(2460753.3393179337), np.float64(2460753.339429782)], [np.float64(2460753.339765326), np.float64(2460753.3398771742)], [np.float64(2460753.3399890224), np.float64(2460753.3401008705)], [np.float64(2460753.343568162), np.float64(2460753.3439037064)], [np.float64(2460753.3562069978), np.float64(2460753.356318846)], [np.float64(2460753.357213631), np.float64(2460753.357437327)], [np.float64(2460753.361911251), np.float64(2460753.3621349474)], [np.float64(2460753.363700821), np.float64(2460753.3639245173)], [np.float64(2460753.3650429985), np.float64(2460753.365378543)], [np.float64(2460753.3711946444), np.float64(2460753.3713064925)], [np.float64(2460753.386629683), np.float64(2460753.386853379)], [np.float64(2460753.391774696), np.float64(2460753.391998392)], [np.float64(2460753.39211024), np.float64(2460753.3922220883)], [np.float64(2460753.394682747), np.float64(2460753.394906443)], [np.float64(2460753.3964723162), np.float64(2460753.3968078606)], [np.float64(2460753.397367101), np.float64(2460753.3975907974)], [np.float64(2460753.399939608), np.float64(2460753.4003870003)], [np.float64(2460753.404301684), np.float64(2460753.404413532)], [np.float64(2460753.404637228), np.float64(2460753.4049727726)], [np.float64(2460753.4228684697), np.float64(2460753.423315862)], [np.float64(2460753.4276779382), np.float64(2460753.4277897864)], [np.float64(2460753.4341651285), np.float64(2460753.4343888247)], [np.float64(2460753.4346125205), np.float64(2460753.434948065)], [np.float64(2460753.435171761), np.float64(2460753.4355073054)], [np.float64(2460753.4376324196), np.float64(2460753.437856116)], [np.float64(2460753.4395338376), np.float64(2460753.4396456857)], [np.float64(2460753.453626699), np.float64(2460753.453850395)], [np.float64(2460753.4542977875), np.float64(2460753.454857028)], [np.float64(2460753.4606731297), np.float64(2460753.460896826)], [np.float64(2460753.468055105), np.float64(2460753.468166953)], [np.float64(2460753.468726193), np.float64(2460753.468838041)], [np.float64(2460753.4689498893), np.float64(2460753.4690617374)], [np.float64(2460753.470515763), np.float64(2460753.470739459)], [np.float64(2460753.4712986997), np.float64(2460753.471410548)], [np.float64(2460753.472529029), np.float64(2460753.4729764215)], [np.float64(2460753.4773384975), np.float64(2460753.4775621938)], [np.float64(2460753.4818124217), np.float64(2460753.48192427)], [np.float64(2460753.4847204727), np.float64(2460753.485279713)], [np.float64(2460753.4867337383), np.float64(2460753.487181131)], [np.float64(2460753.4906484224), np.float64(2460753.4907602705)], [np.float64(2460753.5003792075), np.float64(2460753.5004910557)], [np.float64(2460753.500602904), np.float64(2460753.5010502962)], [np.float64(2460753.5020569293), np.float64(2460753.502280625)], [np.float64(2460753.506866398), np.float64(2460753.507201942)], [np.float64(2460753.5111166257), np.float64(2460753.511228474)], [np.float64(2460753.5125706512), np.float64(2460753.5127943475)], [np.float64(2460753.513577284), np.float64(2460753.513912828)], [np.float64(2460753.5151431574), np.float64(2460753.515702398)], [np.float64(2460753.5161497905), np.float64(2460753.516485335)], [np.float64(2460753.5171564235), np.float64(2460753.517491968)], [np.float64(2460753.518498601), np.float64(2460753.518610449)], [np.float64(2460753.5205118665), np.float64(2460753.5214066515)], [np.float64(2460753.5242028544), np.float64(2460753.5243147025)], [np.float64(2460753.526104272), np.float64(2460753.526439816)], [np.float64(2460753.528117538), np.float64(2460753.5284530823)], [np.float64(2460753.5321440697), np.float64(2460753.532479614)], [np.float64(2460753.5327033103), np.float64(2460753.5329270065)], [np.float64(2460753.534269184), np.float64(2460753.53449288)], [np.float64(2460753.5370653863), np.float64(2460753.5371772344)], [np.float64(2460753.53863126), np.float64(2460753.538854956)], [np.float64(2460753.54042083), np.float64(2460753.5408682222)], [np.float64(2460753.543217032), np.float64(2460753.5434407284)], [np.float64(2460753.5465724757), np.float64(2460753.54690802)], [np.float64(2460753.5516056404), np.float64(2460753.5517174886)], [np.float64(2460753.5518293367), np.float64(2460753.552164881)], [np.float64(2460753.565922198), np.float64(2460753.566034046)], [np.float64(2460753.5748700467), np.float64(2460753.574981895)], [np.float64(2460753.5761003755), np.float64(2460753.5763240717)], [np.float64(2460753.5780017935), np.float64(2460753.578337338)], [np.float64(2460753.579455819), np.float64(2460753.5799032114)], [np.float64(2460753.58292311), np.float64(2460753.5831468063)], [np.float64(2460753.5885155154), np.float64(2460753.5887392117)], [np.float64(2460753.595450098), np.float64(2460753.5956737944)], [np.float64(2460753.599588478), np.float64(2460753.599700326)], [np.float64(2460753.6001477186), np.float64(2460753.622629188)], [np.float64(2460753.622741036), np.float64(2460753.6228528842)], [np.float64(2460753.6231884286), np.float64(2460753.6233002767)], [np.float64(2460753.623523973), np.float64(2460753.623635821)], [np.float64(2460753.6238595173), np.float64(2460753.6239713654)], [np.float64(2460753.624530606), np.float64(2460753.624642454)], [np.float64(2460753.6248661503), np.float64(2460753.6249779984)], [np.float64(2460753.631017796), np.float64(2460753.6311296443)], [np.float64(2460753.635156176), np.float64(2460753.6354917204)], [np.float64(2460753.6367220497), np.float64(2460753.636833898)], [np.float64(2460753.6371694417), np.float64(2460753.63728129)], [np.float64(2460753.637393138), np.float64(2460753.637504986)], [np.float64(2460753.640524885), np.float64(2460753.6407485814)], [np.float64(2460753.6430973914), np.float64(2460753.6432092395)], [np.float64(2460753.6444395687), np.float64(2460753.645446202)], [np.float64(2460753.64555805), np.float64(2460753.645669898)], [np.float64(2460753.6462291386), np.float64(2460753.6463409867)], [np.float64(2460753.646452835), np.float64(2460753.6663617976)]] freq_flags: [[np.float64(47103881.8359375), np.float64(47225952.1484375)], [np.float64(47470092.7734375), np.float64(47592163.0859375)], [np.float64(49911499.0234375), np.float64(50155639.6484375)], [np.float64(62118530.2734375), np.float64(62850952.1484375)], [np.float64(69931030.2734375), np.float64(70053100.5859375)], [np.float64(73959350.5859375), np.float64(75180053.7109375)], [np.float64(77499389.6484375), np.float64(78598022.4609375)], [np.float64(87265014.6484375), np.float64(108261108.3984375)], [np.float64(109970092.7734375), np.float64(110092163.0859375)], [np.float64(112167358.3984375), np.float64(112411499.0234375)], [np.float64(112533569.3359375), np.float64(113143920.8984375)], [np.float64(113265991.2109375), np.float64(113388061.5234375)], [np.float64(113632202.1484375), np.float64(113754272.4609375)], [np.float64(116073608.3984375), np.float64(116195678.7109375)], [np.float64(116439819.3359375), np.float64(116561889.6484375)], [np.float64(116683959.9609375), np.float64(116806030.2734375)], [np.float64(124740600.5859375), np.float64(125228881.8359375)], [np.float64(127548217.7734375), np.float64(127670288.0859375)], [np.float64(129989624.0234375), np.float64(130111694.3359375)], [np.float64(136337280.2734375), np.float64(136459350.5859375)], [np.float64(136947631.8359375), np.float64(138046264.6484375)], [np.float64(138656616.2109375), np.float64(138778686.5234375)], [np.float64(139511108.3984375), np.float64(139633178.7109375)], [np.float64(141464233.3984375), np.float64(141586303.7109375)], [np.float64(141708374.0234375), np.float64(141830444.3359375)], [np.float64(142074584.9609375), np.float64(142318725.5859375)], [np.float64(142440795.8984375), np.float64(142562866.2109375)], [np.float64(142684936.5234375), np.float64(143539428.7109375)], [np.float64(143783569.3359375), np.float64(144027709.9609375)], [np.float64(145492553.7109375), np.float64(145736694.3359375)], [np.float64(146224975.5859375), np.float64(146347045.8984375)], [np.float64(147445678.7109375), np.float64(147567749.0234375)], [np.float64(149154663.0859375), np.float64(149276733.3984375)], [np.float64(149887084.9609375), np.float64(150009155.2734375)], [np.float64(153060913.0859375), np.float64(153182983.3984375)], [np.float64(153427124.0234375), np.float64(153549194.3359375)], [np.float64(154159545.8984375), np.float64(154403686.5234375)], [np.float64(155014038.0859375), np.float64(155136108.3984375)], [np.float64(155258178.7109375), np.float64(155380249.0234375)], [np.float64(155502319.3359375), np.float64(155624389.6484375)], [np.float64(156845092.7734375), np.float64(156967163.0859375)], [np.float64(157577514.6484375), np.float64(157699584.9609375)], [np.float64(158187866.2109375), np.float64(158309936.5234375)], [np.float64(158432006.8359375), np.float64(158676147.4609375)], [np.float64(159164428.7109375), np.float64(159286499.0234375)], [np.float64(159896850.5859375), np.float64(160385131.8359375)], [np.float64(163681030.2734375), np.float64(163803100.5859375)], [np.float64(169906616.2109375), np.float64(170639038.0859375)], [np.float64(170883178.7109375), np.float64(171005249.0234375)], [np.float64(171127319.3359375), np.float64(172103881.8359375)], [np.float64(174911499.0234375), np.float64(175033569.3359375)], [np.float64(175155639.6484375), np.float64(175399780.2734375)], [np.float64(175521850.5859375), np.float64(175765991.2109375)], [np.float64(179672241.2109375), np.float64(179794311.5234375)], [np.float64(181137084.9609375), np.float64(181381225.5859375)], [np.float64(183212280.2734375), np.float64(183334350.5859375)], [np.float64(186386108.3984375), np.float64(186508178.7109375)], [np.float64(187362670.8984375), np.float64(187606811.5234375)], [np.float64(189926147.4609375), np.float64(190048217.7734375)], [np.float64(190780639.6484375), np.float64(191513061.5234375)], [np.float64(191635131.8359375), np.float64(191757202.1484375)], [np.float64(193222045.8984375), np.float64(193344116.2109375)], [np.float64(195663452.1484375), np.float64(195785522.4609375)], [np.float64(197128295.8984375), np.float64(197372436.5234375)], [np.float64(197860717.7734375), np.float64(198715209.9609375)], [np.float64(199203491.2109375), np.float64(199325561.5234375)], [np.float64(200057983.3984375), np.float64(200180053.7109375)], [np.float64(201278686.5234375), np.float64(202133178.7109375)], [np.float64(203231811.5234375), np.float64(203353881.8359375)], [np.float64(204574584.9609375), np.float64(205429077.1484375)], [np.float64(206893920.8984375), np.float64(207015991.2109375)], [np.float64(207138061.5234375), np.float64(207382202.1484375)], [np.float64(208480834.9609375), np.float64(208724975.5859375)], [np.float64(209945678.7109375), np.float64(210067749.0234375)], [np.float64(210433959.9609375), np.float64(210556030.2734375)], [np.float64(211898803.7109375), np.float64(212020874.0234375)], [np.float64(212142944.3359375), np.float64(212631225.5859375)], [np.float64(215194702.1484375), np.float64(215316772.4609375)], [np.float64(220565795.8984375), np.float64(220809936.5234375)], [np.float64(221176147.4609375), np.float64(221298217.7734375)], [np.float64(223129272.4609375), np.float64(223373413.0859375)], [np.float64(225692749.0234375), np.float64(225814819.3359375)], [np.float64(227401733.3984375), np.float64(227523803.7109375)], [np.float64(229110717.7734375), np.float64(229476928.7109375)], [np.float64(229965209.9609375), np.float64(230087280.2734375)], [np.float64(231063842.7734375), np.float64(231185913.0859375)], [np.float64(234237670.8984375), np.float64(234359741.2109375)]] ex_ants: [[np.int64[4], Jee], [np.int64[8], Jee], [np.int64[8], Jnn], [np.int64[10], Jee], [np.int64[10], Jnn], [np.int64[16], Jee], [np.int64[18], Jee], [np.int64[18], Jnn], [np.int64[20], Jnn], [np.int64[21], Jee], [np.int64[22], Jee], [np.int64[22], Jnn], [np.int64[27], Jee], [np.int64[27], Jnn], [np.int64[28], Jee], [np.int64[28], Jnn], [np.int64[29], Jee], [np.int64[29], Jnn], [np.int64[30], Jee], [np.int64[32], Jnn], [np.int64[33], Jnn], [np.int64[34], Jee], [np.int64[34], Jnn], [np.int64[37], Jee], [np.int64[37], Jnn], [np.int64[40], Jnn], [np.int64[42], Jnn], [np.int64[43], Jee], [np.int64[43], Jnn], [np.int64[44], Jee], [np.int64[49], Jnn], [np.int64[51], Jee], [np.int64[55], Jee], [np.int64[57], Jee], [np.int64[57], Jnn], [np.int64[58], Jnn], [np.int64[60], Jee], [np.int64[60], Jnn], [np.int64[64], Jnn], [np.int64[67], Jnn], [np.int64[70], Jee], [np.int64[70], Jnn], [np.int64[71], Jnn], [np.int64[75], Jee], [np.int64[75], Jnn], [np.int64[78], Jee], [np.int64[80], Jnn], [np.int64[81], Jnn], [np.int64[82], Jnn], [np.int64[86], Jee], [np.int64[86], Jnn], [np.int64[87], Jee], [np.int64[92], Jee], [np.int64[97], Jnn], [np.int64[98], Jee], [np.int64[98], Jnn], [np.int64[99], Jnn], [np.int64[102], Jnn], [np.int64[104], Jnn], [np.int64[107], Jee], [np.int64[107], Jnn], [np.int64[108], Jnn], [np.int64[109], Jnn], [np.int64[112], Jee], [np.int64[115], Jee], [np.int64[115], Jnn], [np.int64[117], Jee], [np.int64[120], Jee], [np.int64[121], Jee], [np.int64[130], Jee], [np.int64[130], Jnn], [np.int64[134], Jee], [np.int64[134], Jnn], [np.int64[135], Jee], [np.int64[136], Jnn], [np.int64[137], Jee], [np.int64[137], Jnn], [np.int64[143], Jnn], [np.int64[147], Jee], [np.int64[147], Jnn], [np.int64[148], Jee], [np.int64[148], Jnn], [np.int64[149], Jee], [np.int64[149], Jnn], [np.int64[150], Jee], [np.int64[150], Jnn], [np.int64[154], Jnn], [np.int64[155], Jnn], [np.int64[158], Jnn], [np.int64[161], Jnn], [np.int64[164], Jee], [np.int64[164], Jnn], [np.int64[167], Jee], [np.int64[167], Jnn], [np.int64[168], Jee], [np.int64[168], Jnn], [np.int64[169], Jee], [np.int64[169], Jnn], [np.int64[170], Jee], [np.int64[170], Jnn], [np.int64[171], Jnn], [np.int64[176], Jnn], [np.int64[180], Jee], [np.int64[180], Jnn], [np.int64[187], Jnn], [np.int64[188], Jee], [np.int64[188], Jnn], [np.int64[189], Jee], [np.int64[189], Jnn], [np.int64[190], Jee], [np.int64[190], Jnn], [np.int64[191], Jee], [np.int64[191], Jnn], [np.int64[195], Jnn], [np.int64[197], Jnn], [np.int64[199], Jee], [np.int64[199], Jnn], [np.int64[200], Jee], [np.int64[200], Jnn], [np.int64[201], Jee], [np.int64[201], Jnn], [np.int64[202], Jee], [np.int64[202], Jnn], [np.int64[204], Jee], [np.int64[206], Jee], [np.int64[208], Jee], [np.int64[209], Jnn], [np.int64[212], Jnn], [np.int64[213], Jee], [np.int64[213], Jnn], [np.int64[214], Jnn], [np.int64[216], Jnn], [np.int64[218], Jee], [np.int64[218], Jnn], [np.int64[221], Jee], [np.int64[232], Jee], [np.int64[233], Jnn], [np.int64[238], Jnn], [np.int64[239], Jee], [np.int64[240], Jnn], [np.int64[246], Jee], [np.int64[250], Jee], [np.int64[251], Jee], [np.int64[253], Jnn], [np.int64[255], Jee], [np.int64[255], Jnn], [np.int64[262], Jee], [np.int64[262], Jnn], [np.int64[266], Jee], [np.int64[266], Jnn], [np.int64[268], Jee], [np.int64[268], Jnn], [np.int64[320], Jee], [np.int64[320], Jnn], [np.int64[321], Jee], [np.int64[321], Jnn], [np.int64[322], Jee], [np.int64[322], Jnn], [np.int64[323], Jee], [np.int64[323], Jnn], [np.int64[324], Jee], [np.int64[324], Jnn], [np.int64[325], Jee], [np.int64[325], Jnn], [np.int64[326], Jee], [np.int64[326], Jnn], [np.int64[327], Jee], [np.int64[327], Jnn], [np.int64[328], Jee], [np.int64[328], Jnn], [np.int64[329], Jee], [np.int64[329], Jnn], [np.int64[331], Jee], [np.int64[331], Jnn], [np.int64[332], Jee], [np.int64[332], Jnn], [np.int64[333], Jee], [np.int64[333], Jnn], [np.int64[336], Jee], [np.int64[336], Jnn], [np.int64[340], Jee], [np.int64[340], Jnn]]
Metadata¶
In [24]:
for repo in ['hera_cal', 'hera_qm', 'hera_filters', 'hera_notebook_templates', 'pyuvdata']:
exec(f'from {repo} import __version__')
print(f'{repo}: {__version__}')
hera_cal: 3.7.1.dev18+g10e9584 hera_qm: 2.2.1.dev2+ga535e9e hera_filters: 0.1.6.dev9+gf165ec1
hera_notebook_templates: 0.1.dev989+gee0995d pyuvdata: 3.1.3
In [25]:
print(f'Finished execution in {(time.time() - tstart) / 60:.2f} minutes.')
Finished execution in 74.31 minutes.