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 120 *.sum.red_avg_zscore.h5 files starting with /mnt/sn1/data1/2460839/zen.2460839.21078.sum.red_avg_zscore.h5.
The lst_array is not self-consistent with the time_array and telescope location. Consider recomputing with the `set_lsts_from_time_array` method.
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 120 *.sum.smooth.calfits files starting with /mnt/sn1/data1/2460839/zen.2460839.21078.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.')
30.949% of waterfall flagged to start. Flagging an additional 0 integrations and 17 channels. Flagging 0 channels previously flagged 25.00% or more. Flagging 0 times previously flagged 10.00% or more.
All-NaN slice encountered All-NaN slice encountered
Flagging an additional 0 integrations and 7 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. 32.340% of waterfall flagged after flagging whole times and channels with median z > 1.0. 33.616% of waterfall flagged after flagging z > 4.0 outliers.
35.155% of waterfall flagged after watershed flagging on z > 2.0 neighbors of prior flags. Flagging an additional 0 integrations and 0 channels. Flagging 38 channels previously flagged 25.00% or more. Flagging 16 times previously flagged 10.00% or more. Flagging an additional 0 integrations and 0 channels. Flagging 1 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. 40.823% of waterfall flagged after flagging whole times and channels with average z > 1.0.
Mean of empty slice Mean of empty slice Mean of empty slice Casting complex values to real discards the imaginary part Casting complex values to real discards the imaginary part
41.558% of flagging channels that are 4.0σ outliers after delay filtering the time average.
41.883% 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 120 *.sum.flag_waterfall_round_2.h5 files starting with /mnt/sn1/data1/2460839/zen.2460839.21078.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 = [(int(ant[0]), ant[1]) for ant in sorted(always_flagged_ants)]
dt = np.median(np.diff(times))
out_yml_str = 'JD_flags: ' + str([[float(times[flag_stretch][0] - dt / 2), float(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([[float(freqs[flag_stretch][0] - df / 2), float(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/data1/2460839/2460839_aposteriori_flags.yaml ------------------------------------------------------------------------------ JD_flags: [[2460839.211898971, 2460839.212010819], [2460839.218498009, 2460839.2190572494], [2460839.2220771485, 2460839.2223008447], [2460839.224090414, 2460839.2246496547], [2460839.226774769, 2460839.226886617], [2460839.229235427, 2460839.229459123], [2460839.2295709713, 2460839.2299065157], [2460839.2305776044, 2460839.2306894525], [2460839.2309131487, 2460839.231024997], [2460839.2314723893, 2460839.2315842374], [2460839.236617402, 2460839.2367292503], [2460839.237288491, 2460839.2378477314], [2460839.239413605, 2460839.2399728457], [2460839.241538719, 2460839.241650567]] freq_flags: [[46859741.2109375, 46981811.5234375], [49911499.0234375, 50033569.3359375], [60287475.5859375, 61019897.4609375], [62362670.8984375, 63095092.7734375], [68954467.7734375, 69076538.0859375], [69320678.7109375, 83236694.3359375], [87265014.6484375, 108139038.0859375], [112167358.3984375, 112411499.0234375], [112655639.6484375, 112777709.9609375], [112899780.2734375, 113754272.4609375], [116073608.3984375, 116806030.2734375], [120223999.0234375, 120468139.6484375], [124862670.8984375, 125228881.8359375], [129013061.5234375, 130111694.3359375], [131698608.3984375, 131942749.0234375], [136337280.2734375, 138046264.6484375], [141464233.3984375, 141830444.3359375], [142074584.9609375, 142318725.5859375], [143783569.3359375, 144027709.9609375], [145370483.3984375, 145736694.3359375], [146224975.5859375, 146347045.8984375], [147445678.7109375, 147567749.0234375], [149154663.0859375, 149276733.3984375], [149887084.9609375, 150009155.2734375], [154159545.8984375, 154403686.5234375], [155014038.0859375, 155136108.3984375], [155258178.7109375, 155380249.0234375], [155868530.2734375, 156112670.8984375], [157577514.6484375, 157699584.9609375], [158187866.2109375, 158432006.8359375], [158676147.4609375, 158798217.7734375], [159164428.7109375, 159286499.0234375], [161361694.3359375, 161483764.6484375], [170028686.5234375, 170150756.8359375], [170883178.7109375, 171005249.0234375], [175155639.6484375, 175399780.2734375], [181137084.9609375, 181381225.5859375], [183212280.2734375, 183334350.5859375], [186386108.3984375, 186508178.7109375], [187362670.8984375, 187606811.5234375], [189682006.8359375, 189804077.1484375], [189926147.4609375, 190048217.7734375], [191146850.5859375, 191513061.5234375], [193222045.8984375, 193344116.2109375], [196395874.0234375, 196517944.3359375], [196884155.2734375, 197006225.5859375], [197128295.8984375, 197372436.5234375], [198104858.3984375, 198348999.0234375], [199203491.2109375, 199325561.5234375], [200057983.3984375, 200180053.7109375], [200790405.2734375, 200912475.5859375], [201766967.7734375, 201889038.0859375], [203231811.5234375, 203353881.8359375], [204696655.2734375, 205429077.1484375], [206893920.8984375, 207015991.2109375], [207138061.5234375, 207260131.8359375], [208480834.9609375, 208724975.5859375], [209945678.7109375, 210067749.0234375], [210433959.9609375, 210556030.2734375], [211898803.7109375, 212509155.2734375], [213119506.8359375, 213241577.1484375], [215194702.1484375, 215316772.4609375], [215682983.3984375, 215805053.7109375], [220687866.2109375, 220809936.5234375], [222763061.5234375, 223739624.0234375], [225692749.0234375, 225814819.3359375], [227401733.3984375, 227523803.7109375], [227645874.0234375, 227767944.3359375], [229110717.7734375, 229354858.3984375], [231063842.7734375, 231307983.3984375], [231796264.6484375, 232650756.8359375], [232772827.1484375, 233016967.7734375], [233139038.0859375, 234359741.2109375]] ex_ants: [[4, Jee], [5, Jee], [9, Jnn], [10, Jee], [10, Jnn], [15, Jee], [18, Jee], [18, Jnn], [20, Jee], [21, Jee], [27, Jee], [27, Jnn], [28, Jee], [28, Jnn], [29, Jee], [29, Jnn], [30, Jee], [30, Jnn], [32, Jnn], [33, Jnn], [34, Jee], [35, Jnn], [37, Jnn], [40, Jnn], [44, Jee], [46, Jee], [48, Jee], [49, Jnn], [51, Jee], [51, Jnn], [53, Jee], [53, Jnn], [56, Jee], [60, Jnn], [64, Jnn], [66, Jee], [66, Jnn], [68, Jee], [68, Jnn], [70, Jee], [70, Jnn], [71, Jee], [71, Jnn], [78, Jee], [80, Jnn], [81, Jee], [84, Jnn], [87, Jee], [88, Jee], [88, Jnn], [89, Jee], [89, Jnn], [90, Jee], [90, Jnn], [92, Jee], [97, Jnn], [99, Jee], [99, Jnn], [102, Jnn], [104, Jnn], [107, Jee], [107, Jnn], [108, Jee], [108, Jnn], [109, Jnn], [115, Jee], [115, Jnn], [117, Jee], [118, Jnn], [120, Jee], [120, Jnn], [121, Jee], [127, Jee], [131, Jee], [131, Jnn], [135, Jee], [136, Jnn], [150, Jee], [150, Jnn], [151, Jee], [151, Jnn], [152, Jee], [152, Jnn], [153, Jee], [153, Jnn], [154, Jee], [154, Jnn], [158, Jnn], [159, Jee], [161, Jnn], [165, Jee], [166, Jee], [166, Jnn], [167, Jee], [167, Jnn], [170, Jee], [171, Jee], [171, Jnn], [172, Jee], [172, Jnn], [174, Jnn], [175, Jnn], [180, Jee], [180, Jnn], [182, Jee], [184, Jee], [184, Jnn], [185, Jee], [185, Jnn], [186, Jee], [186, Jnn], [188, Jee], [188, Jnn], [191, Jnn], [194, Jnn], [195, Jnn], [197, Jnn], [199, Jnn], [200, Jee], [200, Jnn], [201, Jee], [201, Jnn], [202, Jee], [202, Jnn], [206, Jnn], [208, Jnn], [212, Jnn], [213, Jee], [214, Jnn], [218, Jnn], [220, Jee], [220, Jnn], [221, Jee], [221, Jnn], [222, Jee], [222, Jnn], [225, Jee], [225, Jnn], [226, Jee], [226, Jnn], [227, Jee], [227, Jnn], [233, Jnn], [236, Jee], [236, Jnn], [237, Jee], [237, Jnn], [238, Jee], [238, Jnn], [239, Jee], [239, Jnn], [240, Jee], [240, Jnn], [250, Jee], [251, Jee], [251, Jnn], [253, Jnn], [254, Jee], [254, Jnn], [255, Jee], [255, Jnn], [256, Jee], [257, Jee], [257, Jnn], [271, Jee], [271, Jnn], [272, Jee], [272, Jnn], [273, Jee], [273, Jnn], [284, Jee], [286, Jnn], [300, Jnn], [301, Jee], [302, Jee], [311, Jee], [311, Jnn], [312, Jee], [312, Jnn], [313, Jee], [314, Jee], [320, Jee], [320, Jnn], [321, Jee], [321, Jnn], [323, Jee], [323, Jnn], [324, Jee], [324, Jnn], [325, Jee], [325, Jnn], [326, Jee], [326, Jnn], [327, Jee], [327, Jnn], [328, Jee], [328, Jnn], [329, Jee], [329, Jnn], [331, Jee], [331, Jnn], [332, Jee], [332, Jnn], [333, Jee], [333, Jnn], [336, Jee], [336, Jnn], [340, Jee], [340, Jnn], [342, Jee], [342, Jnn], [343, Jee], [343, Jnn], [346, Jee], [346, Jnn], [347, Jee], [347, 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.dev45+g4a0c6f1 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 2.55 minutes.