-
Notifications
You must be signed in to change notification settings - Fork 10
/
read_idb.py
1340 lines (1272 loc) · 58.2 KB
/
read_idb.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
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
''' Main routine to read the auto- and cross-correlation data from IDB files.
'''
#
# 2016-03-30 DG
# Began writing the code, adapted from get_X_data2.py
# 2016-04-01 DG
# Added summary_plot()
# 2016-05-12 DG
# Truncate arrays in readXdata() when the file ends early
# 2016-05-14 DG
# Several changes. Changed filter keyword in readXdata()
# to default to False, since we are missing a lot of
# frequencies. Instead any filtering will be done in
# read_idb() after arrays are concatenated. Also added
# time-averaging in read_idb(). Also added flag_sk(),
# which eliminates bad data due to RFI (sort of)
# 2016-06-23 BC
# Changed the default search directory of the EOVSA IDB files
# in get_trange_files() from '/dppdata1/IDB' to /data1/eovsa/fits/IDB'
# 2016-06-30 DG
# Change to allow input of a file list to read_idb() in place of
# trange (used by the realtime pipeline).
# 2016-07-23 DG
# Widen range of allowed SK to 0.7 - 1.5.
# 2016-11-19 DG
# Fix glitches in uvw when averaging over time, by replacing
# missing values with nan before averaging. Also added srcchk boolean
# to read_idb(), to skip the name check if False. Also added the
# option to specify a source name, and changed behavior so that
# if a different source name is found, it just skips the file
# instead of bailing out.
# 2016-12-28 DG
# Fixed error in wrapping of HA. Added new "production" routine
# unrot_miriad(), which does the correction of raw IDB files for
# differential feed rotation and writes out a new file. This
# routine probably needs to move to another module, but it is
# still undergoing debugging.
# 2017-Jan-05 BC, DG
# Added "reverse" parameter to unrot() so that simulation can go
# either unrotating (reverse = False, default), or forward rotating
# (reverse = True). Also remove hard-coded 500 in xsampler and
# ysampler reading, now that IDB data no longer have non-existent
# frequencies.
# 2017-Jan-11 DG
# Added effect of multi-band delay to both unrot_miriad() [Jones
# matrix approach] and unrot_miriad2() [Mueller matrix approach].
# 2017-Jan-19 DG
# Change unrot_miriad() to use difference in feed angle on each
# baseline rather than the angles themselves. Also implements a
# multi-band delay on each antenna.
# 2017-Jan-25 DG
# Further changes to unrot_miriad(), and removed unrot_miriad2().
# Also added code to handle extraneous nulls ('\x00') at end of
# string vars written by aipy.
# 2017-Jan-27 DG
# Fixed some bugs associated with tp_only.
# 2017-Feb-01 DG
# Added read_udb() routine.
# 2017-Apr-04 DG
# Added read_npz() routine to read and concatenate multiple NPZ files.
# 2017-Apr-14 DG
# Very occasional 0-filled record in IDB file was throwing
# off summed times when navg was set in read_idb(). Now readXdata()
# detects this and skips that record (a 1-s data gap).
# 2017-Apr-27 DG
# Added allday_udb() routine, to read all UDB files for a given
# day and optionally make a nice overview plot of the data.
# 2017-May-02 DG
# Added saving of allday_udb() figure if requested by savfig keyword.
# 2017-May-13 BC
# Added summary_plot_pcal() to plot only baselines correlating with Ant 14
# Added "quackint" parameter in read_idb() to get rid of first quackint seconds of each file
# Truncate out['uvw'] to have the same shape of the time axis as out['time']
# 2017-May-14 BC
# Added a key (out['band']) in the output of readXdata() to indicate the band name of a specific frequency
# in fghz
# 2017-Jun-12 DG
# Fix allday_udb() to find files on following date up to 09 UT
# 2017-Jul-12
# Fixed long-standing bug in get_trange_files(), which no longer worked on DPP.
# 2017-Jul-13 DG
# Major update to allday_udb() to overplot GOES data (if available) and scan type info.
# 2017-Jul-15 DG
# Simpler calculation for band names from frequency, instead of calling freq2bdname(),
# which anyway does not work...
# 2017-Aug-09 DG
# Change allday_udb() output plot to have fixed timerange, 13:30 - 02:30 UT. Also
# removed the import of spectrogram_fit and pcapture2, which load a lot of stuff for
# very little purpose. The pcapture2 routines bl2ord and ant_str2list were moved
# to util.py. I just commented out the silly selfcal routine that is never used.
# 2017-Aug-11 DG
# The allday_udb() plot was crashing due to GOES-15 data being all zero, so add
# handling of that situation. Also worked on the plot labeling a bit.
# 2018-Mar-18 DG
# Attempt to make read_idb skip incompatible files (not matching shape of first file).
# 2019-Feb-20 DG
# Fix long-standing error mode, where an empty datalist in read_idb() would cause a
# crash. Now returns an empty dictionary.
# 2019-Jun-22 DG
# Fix bandname list to work with either 34 or 52 band data, depending on date. Dates
# prior to 2019-Feb-22 use 34 bands. Also, added allow_pickle=True in np.load() call
# in read_npz(). Looks like a new security "feature" of Python.
# 2020-Jan-25 DG
# Suppress warnings about imaginary total power data in a file, after the first one.
# 2020-05-10 DG
# Updated cal_qual() to use util.get_idbdir() to find IDB root path.
# 2020-05-11 DG
# Further update to make this work on the DPP.
# 2020-05-29 SY
# Fix bug with datadir.find('eovsa') in case of using /nas3/IDB/
# 2020-12-13 DG
# Important change to honor flags in the IDB files by setting flagged data to NaN.
# 2020-12-16 DG
# Commented out irrelevant (and potentially dangerous, because out of date) read_udb()
# routine. The standard read_idb() works fine for UDB files.
# 2021-01-10 DG
# Important changes to add desat parameter to readXdata (and read_idb), to call the
# new routine autocorr_desat() to correct for correlator saturation. Default is False
# right now, but will likely be changed to True after some tests.
# 2021-01-12 DG
# I found a better way to determine the correction factor using the SK m values to
# adjust for variations in power level due to different numbers of subchannels.
# Also added a test for date that allows this to work for older and newer data.
# 2021-06-01 DG
# It seems the IDB files occasionally have time glitches where the time jumps back,
# I think due to resetting the network every 15 minutes. Now readXdata() just skips
# such records.
# 2022-01-30 DG
# Found a bug in read_idb(), where the frequency list was from the last file read,
# even if it did not match the first file.
# 2022-03-15 DG
# Added an nmax parameter to read_idb() and readXdata() to override previous limitation
# of reading only 600 times from a file. With the new 20-ms files there can be 30000
# records in a 10-min file!
# 2024-03-29 DG
# Change unrot() to agree with the one in pipeline_cal.py.
#
import aipy
import os
from util import Time, nearest_val_idx, bl2ord, ant_str2list, common_val_idx, lobe, get_idbdir, freq2bdname
import glob
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.dates import DateFormatter
#import spectrogram_fit as sp
#import pcapture2 as p
import eovsa_lst as el
import copy
#import chan_util_bc as cu
#import chan_util_52 as cu52
#bl2ord = p.bl_list()
# def read_udb(filename):
# ''' This routine reads the data from a UDB file.
# '''
# # Open uv file for reading
# uv = aipy.miriad.UV(filename)
# nf = len(uv['sfreq'])
# nt = uv['ntimes']
# print_warning = True # Print a warning about imaginary total power data, if found.
# freq = uv['sfreq']
# npol = uv['npol']
# nants = uv['nants']
# nbl = nants*(nants-1)/2
# outa = np.zeros((nants,npol,nf,nt),dtype=np.complex64) # Auto-correlations
# outx = np.zeros((nbl,npol,nf,nt),dtype=np.complex64) # Cross-correlations
# #outp = np.zeros((nants,2,nf,600),dtype=np.float)
# #outp2 = np.zeros((nants,2,nf,600),dtype=np.float)
# #outm = np.zeros((nants,2,nf,600),dtype=np.int)
# uvwarray = np.zeros((nbl,nt,3),dtype=np.float)
# timearray = []
# lstarray = []
# l = -1
# tprev = 0
# tsav = 0
# # Use antennalist if available
# ants = uv['antlist']
# while ants[-1] == '\x00': ants = ants[:-1]
# antlist = map(int, ants.split())
# src = uv['source']
# while src[-1] == '\x00': src = src[:-1]
# for preamble, data in uv.all():
# uvw, t, (i0,j0) = preamble
# i = antlist.index(i0+1)
# j = antlist.index(j0+1)
# if i > j:
# # Reverse order of indices
# j = antlist.index(i0+1)
# i = antlist.index(j0+1)
# # Assumes uv['pol'] is one of -5, -6, -7, -8
# k = -5 - uv['pol']
# if t != tprev:
# # New time
# l += 1
# if l == nt:
# break
# tprev = t
# timearray.append(t)
# #xdata = uv['xsampler'].reshape(nf_orig,nants,3)
# #ydata = uv['ysampler'].reshape(nf_orig,nants,3)
# #outp[:,0,:,l] = np.swapaxes(xdata[:,:,0],0,1)
# #outp[:,1,:,l] = np.swapaxes(ydata[:,:,0],0,1)
# #outp2[:,0,:,l] = np.swapaxes(xdata[:,:,1],0,1)
# #outp2[:,1,:,l] = np.swapaxes(ydata[:,:,1],0,1)
# #outm[:,0,:,l] = np.swapaxes(xdata[:,:,2],0,1)
# #outm[:,1,:,l] = np.swapaxes(ydata[:,:,2],0,1)
# if i0 == j0:
# # This is an auto-correlation
# outa[i0,k,:,l] = data
# if print_warning and k < 2 and np.sum(data != np.real(data)) > 0:
# print preamble,uv['pol'], 'has imaginary data! Additional warnings suppressed.'
# print_warning = False
# else:
# outx[bl2ord[i,j],k,:,l] = data
# if k == 3: uvwarray[bl2ord[i,j],l] = uvw
# # Truncate in case of early end of data
# #nt = len(timearray)
# #outp = outp[:,:,:,:nt]
# #outp2 = outp2[:,:,:,:nt]
# #outm = outm[:,:,:,:nt]
# #if not tp_only:
# # outa = outa[:,:,:,:nt]
# # outx = outx[:,:,:,:nt]
# if len(lstarray) != 0:
# pass
# else:
# tarray = Time(timearray,format='jd')
# for t in tarray:
# lstarray.append(el.eovsa_lst(t))
# ha = np.array(lstarray) - uv['ra']
# ha[np.where(ha > np.pi)] -= 2*np.pi
# ha[np.where(ha < -np.pi)] += 2*np.pi
# out = {'a':outa, 'x':outx, 'uvw':uvwarray, 'fghz':freq, 'time':np.array(timearray),'source':src,'ha':ha,'ra':uv['ra'],'dec':uv['dec']}#,'p':outp,'p2':outp2,'m':outm
# return out
def readXdata(filename, filter=False, tp_only=False, src=None, desat=False, nmax=600):
''' This routine reads the data from a single IDBfile.
Optional Keywords:
filter boolean--if True, returns only non-zero frequencies
if False (default), returns uniform set of 500 frequencies
tp_only boolean--if True, returns only TP information
if False (default), returns everything (including
auto & cross correlations)
nmax max number of times to read from the file. Defaults to 600,
which is the number of records in a 10-min file of 1-s
records. This affects memory usage if it is much larger
than the number of times in the file, but with the new
20-ms files there can be 30000 recs in a 10-min file.
'''
# Open uv file for reading
uv = aipy.miriad.UV(filename)
nf_orig = len(uv['sfreq'])
good_idx = np.arange(nf_orig)
if filter:
good_idx = []
# Read a bunch of records to get number of good frequencies, i.e. those with at least
# some non-zero data. Read 20 records for baseline 1-2, XX pol
uv.select('antennae',0,2,include=True)
uv.select('polarization',-5,-5,include=True)
for i in range(20):
preamble, data = uv.read()
idx, = data.nonzero()
if len(idx) > len(good_idx):
good_idx = copy.copy(idx)
uv.select('clear',0,0)
uv.rewind()
if 'source' in uv.vartable:
source = uv['source']
while source[-1] == '\x00': source = source[:-1]
if src is None:
# If no source name is given, return the source from the file and keep going
src = source
elif src != source:
# If a specific source name is given, and it does not match the file, stop and return None
return source
else:
# If a source is given, and it matches the file, keep going
pass
else:
if src:
# If a specific source name is given, and there is no source in the file, stop and return
pass#return '<no "source" var!>'
print_warning = True # Print a warning about imaginary total power data, if found.
nf = len(good_idx)
freq = uv['sfreq'][good_idx]
npol = uv['npol']
nants = uv['nants']
nbl = nants*(nants-1)/2
if not tp_only:
outa = np.zeros((nants,npol,nf,nmax),dtype=np.complex64) # Auto-correlations
outx = np.zeros((nbl,npol,nf,nmax),dtype=np.complex64) # Cross-correlations
outp = np.zeros((nants,2,nf,nmax),dtype=np.float)
outp2 = np.zeros((nants,2,nf,nmax),dtype=np.float)
outm = np.zeros((nants,2,nf,nmax),dtype=np.int)
uvwarray = np.zeros((nbl,nmax,3),dtype=np.float)
timearray = []
lstarray = []
l = -1
tprev = 0
tsav = 0
# Use antennalist if available
if 'antlist' in uv.vartable:
ants = uv['antlist']
while ants[-1] == '\x00': ants = ants[:-1]
antlist = map(int, ants.split())
else:
antlist = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]
for preamble, data in uv.all():
uvw, t, (i0,j0) = preamble
i = antlist.index(i0+1)
j = antlist.index(j0+1)
if i > j:
# Reverse order of indices
j = antlist.index(i0+1)
i = antlist.index(j0+1)
# Assumes uv['pol'] is one of -5, -6, -7, -8
k = -5 - uv['pol']
if filter:
if len(data.nonzero()[0]) == nf:
if t != tprev:
# print preamble
if t == 2440587.5:
# Time is 1970-01-01, which means a zero-filled record, so skip
# the entire thing.
continue
if t < tprev:
# Some kind of glitch, so skip the whole thing
continue
# New time
l += 1
if l == 600:
break
tprev = t
timearray.append(t)
try:
lstarray.append(uv['lst'])
except:
pass
xdata = uv['xsampler'].reshape(nf_orig,nants,3)
ydata = uv['ysampler'].reshape(nf_orig,nants,3)
outp[:,0,:,l] = np.swapaxes(xdata[good_idx,:,0],0,1)
outp[:,1,:,l] = np.swapaxes(ydata[good_idx,:,0],0,1)
outp2[:,0,:,l] = np.swapaxes(xdata[good_idx,:,1],0,1)
outp2[:,1,:,l] = np.swapaxes(ydata[good_idx,:,1],0,1)
outm[:,0,:,l] = np.swapaxes(xdata[good_idx,:,2],0,1)
outm[:,1,:,l] = np.swapaxes(ydata[good_idx,:,2],0,1)
if tp_only:
outa = None
outx = None
else:
if i0 == j0:
# This is an auto-correlation
outa[i0,k,:,l] = data[data.nonzero()]
else:
outx[bl2ord[i,j],k,:,l] = data[data.nonzero()]
if k == 3: uvwarray[bl2ord[i,j],l] = uvw[data.nonzero()]
else:
if t != tprev:
# New time
if t == 2440587.5:
# Time is 1970-01-01, which means a zero-filled record, so skip
# the entire thing.
continue
if t < tprev:
# Some kind of glitch, so skip the whole thing
continue
l += 1
if l == nmax:
break
tprev = t
timearray.append(t)
xdata = uv['xsampler'].reshape(nf_orig,nants,3)
ydata = uv['ysampler'].reshape(nf_orig,nants,3)
outp[:,0,:,l] = np.swapaxes(xdata[:,:,0],0,1)
outp[:,1,:,l] = np.swapaxes(ydata[:,:,0],0,1)
outp2[:,0,:,l] = np.swapaxes(xdata[:,:,1],0,1)
outp2[:,1,:,l] = np.swapaxes(ydata[:,:,1],0,1)
outm[:,0,:,l] = np.swapaxes(xdata[:,:,2],0,1)
outm[:,1,:,l] = np.swapaxes(ydata[:,:,2],0,1)
if tp_only:
outa = None
outx = None
else:
if i0 == j0:
# This is an auto-correlation
outa[i0,k,:,l] = data.filled(fill_value=np.nan+np.nan*1j)
if print_warning and k < 2 and np.sum(data != np.real(data)) > 0:
print preamble,uv['pol'], 'has imaginary data! Additional warnings suppressed.'
print_warning = False
else:
outx[bl2ord[i,j],k,:,l] = data.filled(fill_value=np.nan+np.nan*1j)
if k == 3: uvwarray[bl2ord[i,j],l] = uvw
# Truncate in case of early end of data
nt = len(timearray)
outp = outp[:,:,:,:nt]
outp2 = outp2[:,:,:,:nt]
outm = outm[:,:,:,:nt]
uvwarray = uvwarray[:,:nt]
if not tp_only:
outa = outa[:,:,:,:nt]
outx = outx[:,:,:,:nt]
if len(lstarray) != 0:
pass
else:
tarray = Time(timearray,format='jd')
for t in tarray:
lstarray.append(el.eovsa_lst(t))
ha = np.array(lstarray) - uv['ra']
ha[np.where(ha > np.pi)] -= 2*np.pi
ha[np.where(ha < -np.pi)] += 2*np.pi
# Find out band name for each frequency
bd = freq2bdname(freq, Time(timearray[0],format='jd'))
out = {'a':outa, 'x':outx, 'uvw':uvwarray, 'fghz':freq, 'band':bd,'time':np.array(timearray),'source':src,'p':outp,'p2':outp2,'m':outm,'ha':ha,'ra':uv['ra'],'dec':uv['dec']}
if desat:
out = autocorr_desat(out)
return out
def autocorr_desat(out):
''' Corrects for correlator saturation effects. Applies a correction to
auto- and cross-correlation amplitudes based on total power amplitudes.
Calculates the function eta = (x + d - c)/[(a*erf((x-c)/b) + d], where x = log(P),
a,b,c,d = [1.22552, 1.37369, 2.94536, 2.14838], and erf() is the error function.
However, eta is set to 1 for A < 50.
Applies the function to autocorrelations A_i and cross-correlations xi_ij to obtain
A'_i = A**eta_i and xi'_ij = xi_ij**[(eta_i + eta_j)/2].
'''
from scipy.special import erf
def eta_f(x, A):
''' This is the desaturation fuction for data taken with equalizer coefficient 8.0,
which is data prior to 5/16/2021.
'''
a,b,c,d = [1.22552, 1.37369, 2.94536, 2.14838] # Parameters define the invariant saturation curve
eta = (x + d - c)/(a*erf((x-c)/b) + d)
bad = np.where(A < 50)
eta[bad] = 1.0
return eta
def eta_2(x, A):
''' This is the desaturation fuction for data taken with equalizer coefficient 2.0,
which is data after to 5/16/2021.
'''
a1,b1,c1,d1 = [0.88025122, 1.0221639 , 4.39845723, 2.38911615]
a2,b2,c2,d2 = [2.28517281, 2.64619331, 4.38657476, 2.37753165]
hi = np.where(A > 300)
low = np.where(A <= 300)
eta = np.ones_like(x)
eta[hi] = (x[hi] + d1 - c1)/(a1*erf((x[hi]-c1)/b1) + d1)
eta[low] = (x[low] + d2 - c2)/(a2*erf((x[low]-c2)/b2) + d2)
return eta
nant = 16
# Determine required "m" value for standardized power level. The power changes
# depending on number of channels averaged, etc., and the SK m value keeps track
# of all of that.
mjd = Time(out['time'][0],format='jd').mjd
if mjd > 58536:
m0 = 745472. # Standard for most recent data (325 MHz bandwidth)
else:
m0 = 721536. # Standard for earlier data (ca. 2017)
nf, = out['fghz'].shape
nt, = out['time'].shape
npol = 4
# Copy power for each polarization so modification below does not affect
# actual data
Px = out['p'][:,0]*m0/out['m'][:,0]
Py = out['p'][:,1]*m0/out['m'][:,0]
# n0 = len(np.where(out['band'] == np.max(out['band']))[0])
# # Modify measured power to correct for variable science-channel bandwidth
# for i in range(nf):
# bnd = out['band'][i]
# n = np.float(len(np.where(out['band'] == bnd)[0]))
# Px[:,i] *= n/n0
# Py[:,i] *= n/n0
x = np.log10(Px)
# Calculate correction for X pol (returns size [nant, nf, nt])
if mjd < 59350: # 2021-05-16
eta_x = eta_f(x, abs(out['a'][:,0]))
else:
eta_x = eta_2(x, abs(out['a'][:,0]))
x = np.log10(Py)
# Calculate correction for Y pol (returns size [nant, nf, nt])
if mjd < 59350: # 2021-05-16
eta_y = eta_f(x, abs(out['a'][:,1]))
else:
eta_y = eta_2(x, abs(out['a'][:,1]))
eta = np.zeros_like(out['x'])
# Loop over baselines for cross-correlation and calculate correction for
# different polarization states.
for i in range(nant-1):
for j in range(i+1,nant):
eta[bl2ord[i,j],0] = eta_x[i]+eta_x[j]
eta[bl2ord[i,j],1] = eta_y[i]+eta_y[j]
eta[bl2ord[i,j],2] = eta_x[i]+eta_y[j]
eta[bl2ord[i,j],3] = eta_x[j]+eta_y[i]
eta = eta/2.
# Correction is to log of values, so apply by raising amplitudes to eta power,
# but preserve the phase
amp = abs(out['x'])
pha = np.angle(out['x'])
amp = amp**eta
out['x'] = amp*np.exp(1j*pha)
# Repeat for auto-correlation by looping over antennas
# This "re-uses" eta to overwrite first nant values
for i in range(nant):
eta[i,0] = eta_x[i]
eta[i,1] = eta_y[i]
eta[i,2] = (eta_x[i] + eta_y[i])/2.
eta[i,3] = (eta_x[i] + eta_y[i])/2.
amp = abs(out['a'])
pha = np.angle(out['a'])
amp = amp**eta[:nant]
out['a'] = amp*np.exp(1j*pha)
return out
def readXdatmp(filename):
# This temporary routine reads the data from a single IDBfile where the
# polarization was written incorrectly. (-1,-2,0,0) instead of (-5,-6,-7,-8)
# the IDB array sorts basline pairs into correct order for an output
# array that just has 18 baselines (2 sets of 4 antennas correlated separately
# now just need to convert i,j into slot in 136-slot array: 15*16/2 corrs +16 auto
# this array corresponds to the first index (auto corr) for each antenna
# 16*(iant-1)-(iant-1)(iant-2)/2
ibl = np.array( [ 0,16,31,45,58,70,81,91,100,108,115,121,126,130,133,135 ])
# Open uv file for reading
uv = aipy.miriad.UV(filename)
good_idx = []
# Read a bunch of records to get number of good frequencies, i.e. those with at least
# some non-zero data. Read 20 records for baseline 1-2, XX pol
uv.select('antennae',0,2,include=True)
uv.select('polarization',-1,-1,include=True)
for i in range(20):
preamble, data = uv.read()
idx, = data.nonzero()
if len(idx) > len(good_idx):
good_idx = copy.copy(idx)
if 'source' in uv.vartable:
src = uv['source']
uv.select('clear',0,0)
uv.rewind()
nf = len(good_idx)
freq = uv['sfreq'][good_idx]
npol = uv['npol']
nants = uv['nants']
nbl = nants*(nants-1)/2
outa = np.zeros((nants,npol,nf,600),dtype=np.complex64) # Auto-correlations
outx = np.zeros((nbl,npol,nf,600),dtype=np.complex64) # Cross-correlations
outp = np.zeros((nants,2,nf,600),dtype=np.float)
outp2 = np.zeros((nants,2,nf,600),dtype=np.float)
outm = np.zeros((nants,2,nf,600),dtype=np.int)
uvwarray = []
timearray = []
l = -1
tprev = 0
# Use antennalist if available
if 'antlist' in uv.vartable:
ants = uv['antlist']
antlist = map(int, ants.split())
else:
antlist = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]
kp = 0
for preamble, data in uv.all():
uvw, t, (i0,j0) = preamble
i = antlist.index(i0+1)
j = antlist.index(j0+1)
if i > j:
# Reverse order of indices
j = antlist.index(i0+1)
i = antlist.index(j0+1)
# Assumes uv['pol'] is -1, -2, 0, 0
if i == 0 and j == 0:
if uv['pol'] == -1:
k = 0
elif uv['pol'] == -2:
k = 1
elif uv['pol'] == 0:
if kp == 1:
k = 2
kp = 0
else:
k = 3
kp = 1
if k == 3:
uvwarray.append(uvw)
if len(data.nonzero()[0]) == nf:
if t != tprev:
# New time
l += 1
if l == 600:
break
tprev = t
timearray.append(t)
xdata = uv['xsampler'].reshape(500,nants,3)
ydata = uv['ysampler'].reshape(500,nants,3)
outp[:,0,:,l] = np.swapaxes(xdata[good_idx,:,0],0,1)
outp[:,1,:,l] = np.swapaxes(ydata[good_idx,:,0],0,1)
outp2[:,0,:,l] = np.swapaxes(xdata[good_idx,:,1],0,1)
outp2[:,1,:,l] = np.swapaxes(ydata[good_idx,:,1],0,1)
outm[:,0,:,l] = np.swapaxes(xdata[good_idx,:,2],0,1)
outm[:,1,:,l] = np.swapaxes(ydata[good_idx,:,2],0,1)
if i0 == j0:
# This is an auto-correlation
outa[i0,k,:,l] = data[data.nonzero()]
else:
outx[bl2ord[i,j],k,:,l] = data[data.nonzero()]
out = {'a':outa, 'x':outx, 'uvw':np.array(uvwarray), 'fghz':freq, 'time':np.array(timearray),'source':src,'p':outp,'p2':outp2,'m':outm}
return out
def summary_plot(out,ant_str='ant1-13',ptype='phase',pol='XX-YY'):
''' Makes a summary amplitude or phase plot for all baselines from ants in ant_str
in out dictionary.
'''
import matplotlib.pyplot as plt
ant_list = ant_str2list(ant_str)
nant = len(ant_list)
if ptype != 'amp' and ptype != 'phase':
print "Invalid plot type. Must be 'amp' or 'phase'."
return
poloff = 0
if pol != 'XX-YY':
poloff = 2
f, ax = plt.subplots(nant,nant)
f.subplots_adjust(hspace=0,wspace=0)
for axrow in ax:
for a in axrow:
a.xaxis.set_visible(False)
a.yaxis.set_visible(False)
for i in range(nant-1):
ai = ant_list[i]
for j in range(i+1,nant):
aj = ant_list[j]
if ptype == 'phase':
ax[i,j].imshow(np.angle(out['x'][bl2ord[ai,aj],0+poloff]))
ax[j,i].imshow(np.angle(out['x'][bl2ord[ai,aj],1+poloff]))
elif ptype == 'amp':
ax[i,j].imshow(np.abs(out['x'][bl2ord[ai,aj],0+poloff]))
ax[j,i].imshow(np.abs(out['x'][bl2ord[ai,aj],1+poloff]))
for i in range(nant):
ai = ant_list[i]
ax[i,i].text(0.5,0.5,str(ai+1),ha='center',va='center',transform=ax[i,i].transAxes,fontsize=14)
def summary_plot_pcal(out,ant_str='ant1-14',ptype='phase',pol='XX-YY'):
''' Makes a summary amplitude or phase plot for all baselines from ants in ant_str
in out dictionary.
'''
import matplotlib.pyplot as plt
ha=out['ha']
fghz=out['fghz']
ant_list = ant_str2list(ant_str)
nant = len(ant_list)
if ptype != 'amp' and ptype != 'phase':
print "Invalid plot type. Must be 'amp' or 'phase'."
return
poloff = 0
if pol != 'XX-YY':
poloff = 2
f, ax = plt.subplots(nant-1,2,figsize=(5,8))
#f.subplots_adjust(hspace=0,wspace=0)
for axrow in ax:
for a in axrow:
a.xaxis.set_visible(False)
a.yaxis.set_visible(False)
for i in range(nant-1):
ai = ant_list[i]
if ptype == 'phase':
#ax[i,0].pcolormesh(ha,fghz,np.angle(out['x'][bl2ord[ai,13],0+poloff]))
#ax[i,1].pcolormesh(ha,fghz,np.angle(out['x'][bl2ord[ai,13],1+poloff]))
ax[i,0].imshow(np.angle(out['x'][bl2ord[ai,13],0+poloff]))
ax[i,1].imshow(np.angle(out['x'][bl2ord[ai,13],1+poloff]))
elif ptype == 'amp':
#ax[i,0].pcolormesh(ha,fghz,np.abs(out['x'][bl2ord[ai,13],0+poloff]))
#ax[i,1].pcolormesh(ha,fghz,np.abs(out['x'][bl2ord[ai,13],1+poloff]))
ax[i,0].imshow(np.abs(out['x'][bl2ord[ai,13],0+poloff]))
ax[i,1].imshow(np.abs(out['x'][bl2ord[ai,13],1+poloff]))
ax[i,0].text(-0.1,0.5,str(ai+1),ha='center',va='center',transform=ax[i,0].transAxes,fontsize=14)
polstr=pol.split('-')
for j in range(2):
ax[0,j].text(0.5,1.3,polstr[j],ha='center',va='center',transform=ax[0,j].transAxes,fontsize=14)
def get_goes_data(t=None,sat_num=None):
''' Reads GOES data from https://umbra.nascom.nasa.gov/ repository, for date
and satellite number provided. If sat_num is None, data for all available
satellites are downloaded, with some sanity check used to decide the best.
If the Time() object t is None, data for the day before the current date
are read (since there is a delay of 1 day in availability of the data).
Returns:
goes_t GOES time array in plot_date format
goes_data GOES 1-8 A lightcurve
'''
from sunpy.util.config import get_and_create_download_dir
import shutil
from astropy.io import fits
import urllib2
if t is None:
t = Time(Time.now().mjd - 1,format='mjd')
yr = t.iso[:4]
datstr = t.iso[:10].replace('-','')
if sat_num is None:
f = urllib2.urlopen('https://umbra.nascom.nasa.gov/goes/fits/'+yr)
lines = f.readlines()
sat_num = []
for line in lines:
idx = line.find(datstr)
if idx != -1:
sat_num.append(line[idx-2:idx])
if type(sat_num) is int:
sat_num = [str(sat_num)]
filenames = []
for sat in sat_num:
filename = 'go'+sat+datstr+'.fits'
url = 'https://umbra.nascom.nasa.gov/goes/fits/'+yr+'/'+filename
f = urllib2.urlopen(url)
with open(get_and_create_download_dir()+'/'+filename,'wb') as g:
shutil.copyfileobj(f,g)
filenames.append(get_and_create_download_dir()+'/'+filename)
pmerit = 0
for file in filenames:
gfits = fits.open(file)
data = gfits[2].data['FLUX'][0][:,0]
good, = np.where(data > 1.e-8)
tsecs = gfits[2].data['TIME'][0]
merit = len(good)
date_elements = gfits[0].header['DATE-OBS'].split('/')
if merit > pmerit:
print 'File:',file,'is best'
pmerit = merit
goes_data = data
goes_t = Time(date_elements[2]+'-'+date_elements[1]+'-'+date_elements[0]).plot_date + tsecs/86400.
try:
return goes_t, goes_data
except:
print 'No good GOES data for',datstr
return None, None
def allday_udb(t=None, doplot=True, goes_plot=True, savfig=False, gain_corr=False):
# Plots (and returns) UDB data for an entire day
from sunpy import lightcurve
from sunpy.time import TimeRange
from flare_monitor import flare_monitor
if t is None:
t = Time.now()
# Cannot get a GOES plot unless doplot is True
if goes_plot: doplot = True
date = t.iso[:10].replace('-','')
# Look also at the following day, up to 9 UT
date2 = Time(t.mjd + 1,format='mjd').iso[:10].replace('-','')
year = date[:4]
files = glob.glob('/data1/eovsa/fits/UDB/'+year+'/UDB'+date+'*')
files.sort()
files2 = glob.glob('/data1/eovsa/fits/UDB/'+year+'/UDB'+date2+'0*')
files2.sort()
files = np.concatenate((np.array(files),np.array(files2)))
# Eliminate files starting before 10 UT on date (but not on date2)
for i,file in enumerate(files):
if file[-6] != '0':
break
try:
files = files[i:]
except:
print 'No files found in /data1/eovsa/fits/UDB/ for',date
return {}
out = read_idb(files,src='Sun')
if gain_corr:
import gaincal2 as gc
out = gc.apply_gain_corr(out)
trange = Time(out['time'][[0,-1]], format = 'jd')
fghz = out['fghz']
if doplot:
f, ax = plt.subplots(1,1,figsize=(14,5))
pdata = np.sum(np.sum(np.abs(out['x'][0:11,:]),1),0) # Spectrogram to plot
X = np.sort(pdata.flatten()) # Sorted, flattened array
# Set any time gaps to nan
tdif = out['time'][1:] - out['time'][:-1]
bad, = np.where(tdif > 120./86400) # Time gaps > 2 minutes
pdata[:,bad] = 0
vmax = X[int(len(X)*0.95)] # Clip at 5% of points
im = ax.pcolormesh(Time(out['time'],format='jd').plot_date,out['fghz'],pdata,vmax=vmax)
plt.colorbar(im,ax=ax,label='Amplitude [arb. units]')
ax.xaxis_date()
ax.xaxis.set_major_formatter(DateFormatter("%H:%M"))
ax.set_ylim(fghz[0], fghz[-1])
ax.set_xlabel('Time [UT]')
ax.set_ylabel('Frequency [GHz]')
ax.set_title('EOVSA 1-min Data for '+t.iso[:10])
f.autofmt_xdate(bottom=0.15)
if goes_plot:
# Initially assign GOES times as None
goes_t = None
goes_t2 = None
# Get GOES data for overplotting
goes_tr = TimeRange(trange.iso)
goes_label = [' A',' B',' C',' M',' X']
# The GOES label is placed to start 20 min into the day
goes_label_time = Time(out['time'][[0]], format = 'jd').plot_date + 0.014
rightaxis_label_time = trange[1].plot_date
# Retrieve GOES data for the day, but this only goes to end of UT day
goes_t, goes_data = get_goes_data(trange[0])
if int(trange[1].mjd) != int(trange[0].mjd):
goes_t2, goes_data2 = get_goes_data(trange[1])
if goes_t is None and goes_t2 is None:
ax.text (goes_label_time, 12, 'GOES soft x-ray data missing', color = 'yellow')
else:
if not goes_t is None:
goes_data = 2* (np.log10(goes_data + 1.e-9)) + 26
ax.plot_date(goes_t, goes_data,'-',color='yellow')
ytext = np.median(goes_data) - 1
if not goes_t2 is None:
goes_data2 = 2* (np.log10(goes_data2 + 1.e-9)) + 26
ax.plot_date(goes_t2, goes_data2,'-',color='yellow')
ytext2 = np.median(goes_data2) - 1
if ytext:
ytext = (ytext+ytext2)/2
else:
ytext = ytext2
ax.text (goes_label_time, ytext, 'GOES soft x-ray data', color = 'yellow')
# try:
# goes = lightcurve.GOESLightCurve.create(goes_tr)
# if len(np.where(goes.data['xrsb'] != 0.0)[0]) < 100:
# # Looks like the GOES data are all zero, so just skip it
# ax.text (goes_label_time, 12, 'GOES soft x-ray data missing', color = 'yellow')
# else:
# goes.data['xrsb'] = 2* (np.log10(goes.data['xrsb'] + 1.e-9)) + 26
# ytext = np.median(goes.data['xrsb']) - 1
# ax.text (goes_label_time, ytext, 'GOES soft x-ray data', color = 'yellow')
# goes.data['xrsb'].plot(color = 'yellow')
# except:
# # Looks like the GOES data do not exist, so just skip it
# ax.text (goes_label_time, 12, 'GOES soft x-ray data missing', color = 'yellow')
for k,i in enumerate([10,12,14,16,18]):
ax.text(rightaxis_label_time, i-0.4, goes_label[k], fontsize = '12')
ax.plot_date(rightaxis_label_time + np.array([-0.005,0.0]),[i,i],'-',color='yellow')
# try:
# # If the day goes past 0 UT, get GOES data for the next UT day
# if int(trange[1].mjd) != int(trange[0].mjd):
# goes_tr2 = TimeRange([trange[1].iso[:10], trange[1].iso])
# goesday2 = lightcurve.GOESLightCurve.create(goes_tr2)
# if len(np.where(goesday2.data['xrsb'] != 0.0)[0]) < 100:
# pass
# else:
# goesday2.data['xrsb'] = 2* (np.log10(goesday2.data['xrsb'] + 1.e-9)) + 26
# goesday2.data['xrsb'].plot(color = 'yellow')
# except:
# # Looks like the GOES data do not exist, so just skip it
# pass
pstart = Time(t.iso[:10]+' 13:30').plot_date
prange = [pstart,pstart+13./24]
ax.set_xlim(prange)
ut, fl, projdict = flare_monitor(t)
if fl == []:
print 'Error retrieving data for',t.iso[:10],'from SQL database.'
return
if projdict == {}:
print 'No annotation can be added to plot for',t.iso[:10]
else:
defcolor = '#ff7f0e'
nscans = len(projdict['Timestamp'])
SOS = Time(projdict['Timestamp'],format='lv').plot_date
EOS = Time(projdict['EOS'],format='lv').plot_date
yran = np.array(ax.get_ylim())
for i in range(nscans):
uti = SOS[i]*np.array([1.,1.])
#if uti[0] >= trange[0].plot_date:
ax.plot_date(uti,yran,'g',lw=0.5)
if projdict['Project'][i] == 'NormalObserving' or projdict['Project'][i] == 'Normal Observing':
ax.text(uti[0],yran[1]*0.935,'SUN',fontsize=8, color = defcolor, clip_on=True)
elif projdict['Project'][i] == 'None':
ax.text(uti[0],yran[1]*0.975,'IDLE',fontsize=8, color = defcolor, clip_on=True)
elif projdict['Project'][i][:4] == 'GAIN':
ax.text(uti[0],yran[1]*0.955,'GCAL',fontsize=8, color = defcolor, clip_on=True)
elif projdict['Project'][i] == 'SOLPNTCAL':
ax.text(uti[0],yran[1]*0.955,'TPCAL',fontsize=8, color = defcolor, clip_on=True)
elif projdict['Project'][i] == 'PHASECAL':
ax.text(uti[0],yran[1]*0.955,'PCAL',fontsize=8, color = defcolor, clip_on=True)
else:
ax.text(uti[0],yran[1]*0.975,projdict['Project'][i],fontsize=8, color = defcolor, clip_on=True)
known = ['GAIN','PHAS','SOLP'] # known calibration types (first 4 letters)
for i in range(nscans):
uti = EOS[i]*np.array([1.,1.])
ax.plot_date(uti,yran,'r--',lw=0.5)
uti = np.array([SOS[i],EOS[i]])
if projdict['Project'][i] == 'NormalObserving':
ax.plot_date(uti,yran[1]*np.array([0.93,0.93]),ls='-',marker='None',color='#aaffaa',lw=2,solid_capstyle='butt')
elif projdict['Project'][i][:4] in known:
ax.plot_date(uti,yran[1]*np.array([0.95,0.95]),ls='-',marker='None',color='#aaaaff',lw=2,solid_capstyle='butt')
else:
ax.plot_date(uti,yran[1]*np.array([0.97,0.97]),ls='-',marker='None',color='#ffaaaa',lw=2,solid_capstyle='butt')
if savfig:
plt.savefig('/common/webplots/flaremon/daily/'+date[:4]+'/XSP'+date+'.png',bbox_inches='tight')
return out
# def allday_udb(t=None, doplot=True, savfig=False):
# # Plots (and returns) UDB data for an entire day
# if t is None:
# t = Time.now()
# date = t.iso[:10].replace('-','')
# # Look also at the following day, up to 9 UT
# date2 = Time(t.mjd + 1,format='mjd').iso[:10].replace('-','')
# year = date[:4]
# files = glob.glob('/data1/eovsa/fits/UDB/'+year+'/UDB'+date+'*')
# files.sort()
# files2 = glob.glob('/data1/eovsa/fits/UDB/'+year+'/UDB'+date2+'0*')
# files2.sort()
# files = np.concatenate((np.array(files),np.array(files2)))
# # Eliminate files starting before 10 UT on date (but not on date2)
# for i,file in enumerate(files):
# if file[-6] != '0':
# break
# try:
# files = files[i:]
# except:
# print 'No files found in /data1/eovsa/fits/UDB/ for',date
# return {}
# out = read_idb(files,src='Sun')
# if doplot:
# f, ax = plt.subplots(1,1)
# f.set_size_inches(14,5)
# pdata = np.sum(np.sum(np.abs(out['x'][0:11,:]),1),0) # Spectrogram to plot
# X = np.sort(pdata.flatten()) # Sorted, flattened array
# # Set any time gaps to nan
# tdif = out['time'][1:] - out['time'][:-1]
# bad, = np.where(tdif > 120./86400) # Time gaps > 2 minutes
# pdata[:,bad] = 0
# vmax = X[int(len(X)*0.95)] # Clip at 5% of points
# ax.pcolormesh(Time(out['time'],format='jd').plot_date,out['fghz'],pdata,vmax=vmax)
# ax.xaxis_date()
# ax.xaxis.set_major_formatter(DateFormatter("%H:%M"))
# ax.set_xlabel('Time [UT]')
# ax.set_ylabel('Frequency [GHz]')
# ax.set_title('EOVSA 1-min Data for '+t.iso[:10])
# if savfig:
# plt.savefig('/common/webplots/flaremon/XSP_later.png',bbox_inches='tight')
# return out
def read_idb(trange,navg=None, nmax=600, quackint=0.,filter=True,srcchk=True,src=None,tp_only=False, desat=False):
''' This finds the IDB files within a given time range and concatenates
the times into a single dictionary. If trange is not a Time() object,
assume that it is the list of files to read.
Keywords:
src string--if not None, files in trange will be skipped unless
their source name matches this string.
Optional Keywords:
filter boolean--if True (default), returns only non-zero frequencies
if False, returns uniform set of 500 frequencies, with gaps
srcchk boolean--if True (default), stops reading files when source
name is different from initial source name. If set to
False, only the source name in the first file is returned
tp_only boolean--if True, returns only TP information
if False (default), returns everything (including
auto & cross correlations)
quackint float--first time range (in seconds) to skip in the beginning of
each file. Default is 0., or no quack.
'''
if type(trange) == Time:
files = get_trange_files(trange)
else:
# If input type is not Time, assume that it is the list of files to read
files = trange
datalist = []
for file in files:
#This will skip any files that give us errors.
# The names of the bad or unreadable files will
# be printed.
try:
out = readXdata(file,tp_only=tp_only,src=src, desat=desat, nmax=nmax)
if type(out) is str:
print 'Source name:',out,'does not match requested name:',src+'. Will skip',file
else:
if srcchk and src is None:
# This is the first file, and we care about the source, so set source name
src = out['source']