forked from keflavich/brick_linesurvey
-
Notifications
You must be signed in to change notification settings - Fork 0
/
line_imaging.py
258 lines (222 loc) · 10.6 KB
/
line_imaging.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
"""
Line imaging script. There needs to be a to_image.json file in the directory
this is run in. The to_image.json file is produced by the split_windows.py
script.
You can set the following environmental variables for this script:
CHANCHUNKS=<number>
The chanchunks parameter for tclean. Depending on the version, it may
be acceptable to specify this as -1, or it has to be positive. This is
the number of channels that will be imaged all at once; if this is too
large, the data won't fit into memory and CASA will crash.
BAND_NUMBERS=<band(s)>
Image this/these bands. Can be "3", "6", or "3,6" (no quotes)
LOGFILENAME=<name>
Optional. If specified, the logger will use this filenmae
"""
import json
import os
import numpy as np
import astropy.units as u
from astropy import constants
try:
from tasks import tclean, uvcontsub, impbcor
from taskinit import casalog
from tasks import exportfits
except ImportError:
# futureproofing: CASA 6 imports this way
from casatasks import tclean, uvcontsub, impbcor
from casatasks import casalog
from casatasks import exportfits
from parse_contdotdat import parse_contdotdat, freq_selection_overlap
from metadata_tools import determine_imsize, determine_phasecenter, is_7m, logprint
from imaging_parameters import line_imaging_parameters, selfcal_pars, line_parameters
from taskinit import msmdtool, iatool, mstool
from metadata_tools import effectiveResolutionAtFreq
from getversion import git_date, git_version
msmd = msmdtool()
ia = iatool()
ms = mstool()
if os.getenv('LOGFILENAME'):
casalog.setlogfile(os.path.join(os.getcwd(), os.getenv('LOGFILENAME')))
imaging_root = "imaging_results"
if not os.path.exists(imaging_root):
os.mkdir(imaging_root)
# set the 'chanchunks' parameter globally.
# CASAguides recommend chanchunks=-1, but this resulted in: 2018-09-05 23:16:34 SEVERE tclean::task_tclean:: Exception from task_tclean : Invalid Gridding/FTM Parameter set : Must have at least 1 chanchunk
chanchunks = os.getenv('CHANCHUNKS') or 16
# global default: only do robust 0 for lines
robust = 0
vis = os.getenv('VIS')
if vis is None:
raise ValueError("VIS not specified")
logprint("Running on vis {0}".format(vis))
ms.open(vis)
spwinfo = ms.getspectralwindowinfo()
spw_list = spwinfo.keys()
logprint("SPWs are {0}".format(spw_list))
field = 'BrickMaser'
for spw in spw_list:
if spwinfo[spw]['ChanWidth'] < 0:
reference_frequency = spwinfo[spw]['RefFreq'] - spwinfo[spw]['ChanWidth']*spwinfo[spw]['NumChan']
else:
reference_frequency = spwinfo[spw]['RefFreq']
freqname = int(reference_frequency / 1e9)
logprint("SPW {spw} has rest frequency {reference_frequency} = {freqname}".format(**locals()))
lineimagename = os.path.join(imaging_root,
"BrickMaser_{0}_spw{1}".format(freqname,
spw,))
coosys, racen, deccen = determine_phasecenter(ms=vis, field=field)
phasecenter = "{0} {1}deg {2}deg".format(coosys, racen, deccen)
(dra, ddec, pixscale) = determine_imsize(ms=vis, field=field,
phasecenter=(racen, deccen),
spw=0, pixfraction_of_fwhm=1/4.,
min_pixscale=0.1, # arcsec
)
imsize = [int(dra), int(ddec)]
cellsize = ['{0:0.2f}arcsec'.format(pixscale)] * 2
dirty_tclean_made_residual = False
# calculate the channel width
chanwidths = []
msmd.open(vis)
count_spws = len(msmd.spwsforfield(field))
msmd.close()
chanwidth = np.max([np.abs(
effectiveResolutionAtFreq(vis,
spw='{0}'.format(ii),
freq=reference_frequency*u.Hz,
kms=True).max()) for ii in
spw_list])
chanwidths.append(chanwidth)
chanwidth = np.mean(chanwidths)
logprint("Channel widths were {0}, mean = {1}".format(chanwidths,
chanwidth),
origin="almaimf_line_imaging")
impars = {
'niter': 100000,
'robust': 0.0,
'weighting': 'briggs',
'scales': [0,3,9,],
'gridder': 'standard',
'specmode': 'cube',
'deconvolver': 'multiscale',
'outframe': 'LSRK',
'veltype': 'radio',
#'sidelobethreshold': 2.0,
#'noisethreshold': 5.0,
#'usemask': 'auto-multithresh',
'threshold': '5sigma',
'interactive': False,
'pblimit': 0.2,
'nterms': 1,
'spw': spw,
'field': field,
}
impars['chanchunks'] = chanchunks
impars['imsize'] = imsize
impars['cell'] = cellsize
impars['phasecenter'] = phasecenter
#impars['field'] = [field.encode()]*len(vis)
# start with cube imaging
if not os.path.exists(lineimagename+".image") and not os.path.exists(lineimagename+".residual"):
if os.path.exists(lineimagename+".psf"):
logprint("WARNING: The PSF for {0} exists, but no image exists."
" This likely implies that an ongoing or incomplete "
"imaging run for this file exists. It will not be "
"imaged this time; please check what is happening. "
"(this warning issued /before/ dirty imaging)"
.format(lineimagename),
origin='almaimf_line_imaging')
continue
# first iteration makes a dirty image to estimate the RMS
impars_dirty = impars.copy()
impars_dirty['niter'] = 0
logprint("Dirty imaging parameters are {0}".format(impars_dirty),
origin='almaimf_line_imaging')
tclean(vis=vis,
imagename=lineimagename,
restoringbeam='', # do not use restoringbeam='common'
# it results in bad edge channels dominating the beam
**impars_dirty
)
for suffix in ('image', 'residual', 'model'):
ia.open(lineimagename+"."+suffix)
ia.sethistory(origin='almaimf_line_imaging',
history=["{0}: {1}".format(key, val) for key, val in
impars_dirty.items()])
ia.sethistory(origin='almaimf_line_imaging',
history=["git_version: {0}".format(git_version),
"git_date: {0}".format(git_date)])
ia.close()
if os.path.exists(lineimagename+".image"):
# tclean with niter=0 is not supposed to produce a .image file,
# but if it does (and it appears to have done so on at
# least one run), we still want to clean the cube
dirty_tclean_made_residual = True
elif not os.path.exists(lineimagename+".residual"):
raise ValueError("The residual image is required for further imaging.")
else:
logprint("Found existing files matching {0}".format(lineimagename),
origin='almaimf_line_imaging'
)
if os.path.exists(lineimagename+".psf") and not os.path.exists(lineimagename+".image"):
logprint("WARNING: The PSF for {0} exists, but no image exists."
" This likely implies that an ongoing or incomplete "
"imaging run for this file exists. It will not be "
"imaged this time; please check what is happening."
"(warning issued /after/ dirty imaging)"
.format(lineimagename),
origin='almaimf_line_imaging')
# just skip the rest here - that means no contsub imaging
continue
# the threshold needs to be computed if any imaging is to be done (either contsub or not)
# no .image file is produced, only a residual
logprint("Computing residual image statistics for {0}".format(lineimagename), origin='almaimf_line_imaging')
ia.open(lineimagename+".residual")
stats = ia.statistics(robust=True)
rms = float(stats['medabsdevmed'] * 1.482602218505602)
ia.close()
continue_imaging = False
if 'threshold' in impars:
if 'sigma' in impars['threshold']:
nsigma = int(impars['threshold'].strip('sigma'))
threshold = "{0:0.4f}Jy".format(nsigma*rms) # 3 rms might be OK in practice
logprint("Threshold used = {0} = {2}x{1}".format(threshold, rms, nsigma),
origin='almaimf_line_imaging')
impars['threshold'] = threshold
else:
threshold = impars['threshold']
nsigma = (u.Quantity(threshold) / rms).to(u.Jy).value
logprint("Manual threshold used = {0} = {2}x{1}"
.format(threshold, rms, nsigma),
origin='almaimf_line_imaging')
if u.Quantity(threshold).to(u.Jy).value < stats['max']:
# if the threshold was not reached, keep cleaning
continue_imaging = True
if continue_imaging or dirty_tclean_made_residual or not os.path.exists(lineimagename+".image"):
# continue imaging using a threshold
logprint("Imaging parameters are {0}".format(impars),
origin='almaimf_line_imaging')
tclean(vis=vis,
imagename=lineimagename,
restoringbeam='', # do not use restoringbeam='common'
# it results in bad edge channels dominating the beam
**impars
)
for suffix in ('image', 'residual', 'model'):
ia.open(lineimagename+"."+suffix)
ia.sethistory(origin='almaimf_line_imaging',
history=["{0}: {1}".format(key, val) for key, val in
impars.items()])
ia.sethistory(origin='almaimf_line_imaging',
history=["git_version: {0}".format(git_version),
"git_date: {0}".format(git_date)])
ia.close()
impbcor(imagename=lineimagename+'.image',
pbimage=lineimagename+'.pb',
outfile=lineimagename+'.image.pbcor', overwrite=True)
exportfits(lineimagename+".image", lineimagename+".image.fits",
overwrite=True)
exportfits(lineimagename+".image.pbcor", lineimagename+".image.pbcor.fits",
overwrite=True)
logprint("Completed {0}".format(vis), origin='almaimf_line_imaging')
logprint("Completed line_imaging.py run", origin='almaimf_line_imaging')