-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathinterface.py
1059 lines (878 loc) · 44.2 KB
/
interface.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
import logging
import numpy as np
import pandas as pd
import xarray as xr
from pathlib import Path
from netCDF4 import Dataset, date2num
from collections import OrderedDict
from pyldas.grids import EASE2
from pyldas.templates import get_template
from pyldas.paths import paths
logging.basicConfig(format='%(levelname)s: %(message)s', level=logging.INFO)
def s(line):
"""" Remove quotes from string """
return line[1:-2]
def b(line):
"""" Turn 'T' and 'F' into True and False, respectively"""
return True if line[-2] == 'T' else False
class LDAS_io(object):
"""
Class for reading and writing LDAS specific data
Default paths are taken from pyldas.paths
Parameters
----------
param : str
Name of the parameter for which data should be read
exp : string
experiment name (appended to root path)
domain : string
domain name (appended to experiment path)
root : pathlib.Path
root path to the experiment directory
Attributes
----------
paths : dict
Dictionary holding the path information for the specified exp/domain
obsparam : pd.DataFrame
Metadata about observations
param : str
Name of the parameter for which data is loaded
files : np.array
Array containing all names within the specified experiment directory, that match
the specified parameter (excluding netcdf files, if already created)
dates : pd.DatetimeIndex
Dates corresponding to the files in self.files
images : xr.Dataset
netCDF image-chunked image stack (if available)
timeseries : xr.Dataset
netCDF timeseries-chunked image stack (if available)
"""
def __init__(self,
mode,
param=None,
exp=None,
domain=None,
root=None):
self.mode = mode
self.paths = paths(self.mode, exp=exp, domain=domain, root=root)
try:
self.obsparam = self.read_obsparam()
except:
logging.info('No obsparam file. This is a model-only (No DA) run.')
tilecoord = self.read_params('tilecoord')
tilegrids = self.read_params('tilegrids')
self.grid = EASE2(tilecoord, tilegrids)
self.param = param
if param is not None:
if param == 'xhourly':
path = self.paths.cat
else:
path = self.paths.exp_root
nc_file = list(path.glob(f'**/*{param}_images.nc'))
if len(nc_file) == 0:
logging.warning('NetCDF image cube not yet created. Use method "bin2netcdf".')
else:
self.images = xr.open_dataset(nc_file[0])
nc_file = list(path.glob(f'**/*{param}_timeseries.nc'))
if len(nc_file) == 0:
logging.warning('NetCDF time series cube not yet created. Use the NetCDF kitchen sink.')
else:
self.timeseries = xr.open_dataset(nc_file[0])
if param == 'ObsFcstAna':
self.files = np.sort(list(path.glob(f'**/*{param}.*.bin')))
else:
if self.mode == 'GEOSldas':
self.files = np.sort(list(path.glob(f'**/*{param}.*.nc4')))
else:
self.files = np.sort(list(path.glob(f'**/*{param}*.bin')))
if param == 'hscale':
self.pentads = np.array([f.name[-6:-4] for f in self.files]).astype('int')
self.orbits = np.array([f.name[-9:-8] for f in self.files])
else:
if 'monthly' in param:
self.dates = pd.to_datetime([f.name[-10:-4] for f in self.files], format='%Y%m')
else:
self.dates = pd.to_datetime([f.name[-18:-5] for f in self.files], format='%Y%m%d_%H%M')
if param == 'ObsFcstAnaEns':
self.ens_id = np.array([f.name[-42:-38] for f in self.files]).astype('int')
# TODO: Currently valid for 3-hourly data only! Times of the END of the 3hr periods are assigned!
# if self.param == 'xhourly':
# self.dates += pd.to_timedelta('2 hours')
if len(self.files) > 0:
if '.bin' in str(self.files[0]):
self.dtype, self.hdr, self.length = get_template(self.param)
def read_obsparam(self):
""" Read the 'obsparam' file. """
fp = open(list(self.paths.rc_out.glob('**/*obsparam*'))[0])
lines = fp.readlines()[1::]
n_lines = len(lines)
# 30 or 32 fields (before and after two entries for the use of uncertainty maps)
# 34 fields for GEOSldas including the use of uncertainty maps
if n_lines == 128 or n_lines == 896:
n_fields = 32
elif n_lines == 136 or n_lines == 272:
n_fields = 34
else:
n_fields = 30
n_blocks = n_lines / n_fields
# different output scenarios.
res = []
for bl in np.arange(n_blocks, dtype='int') * n_fields:
if n_fields == 32 and n_blocks == 28:
res.append({'descr': s(lines[bl + 0]),
'species': int(lines[bl + 1]),
'orbit': int(lines[bl + 2]),
'pol': int(lines[bl + 3]),
'N_ang': int(lines[bl + 4]),
'ang': float(lines[bl + 5]),
'freq': float(lines[bl + 6]),
'FOV': float(lines[bl + 7]),
'FOV_units': s(lines[bl + 8]),
'assim': b(lines[bl + 9]),
'scale': b(lines[bl + 10]),
'getinnov': b(lines[bl + 11]),
'RTM_ID': int(lines[bl + 12]),
'bias_Npar': int(lines[bl + 13]),
'bias_trel': int(lines[bl + 14]),
'bias_tcut': int(lines[bl + 15]),
'nodata': float(lines[bl + 16]),
'varname': s(lines[bl + 17]),
'units': s(lines[bl + 18]),
'path': s(lines[bl + 19]),
'name': s(lines[bl + 20]),
'scalepath': s(lines[bl + 21]),
'scalename': s(lines[bl + 22]),
'flistpath': s(lines[bl + 23]),
'flistname': s(lines[bl + 24]),
'errstd': float(lines[bl + 25]),
'std_normal_max': float(lines[bl + 26]),
'zeromean': b(lines[bl + 27]),
'coarsen_pert': b(lines[bl + 28]),
'xcorr': float(lines[bl + 29]),
'ycorr': float(lines[bl + 30]),
'adapt': int(lines[bl + 31])})
elif n_fields == 32 and n_blocks == 4:
if self.mode == 'GEOSldas':
res.append({'descr': s(lines[bl + 0]),
'species': int(lines[bl + 1]),
'orbit': int(lines[bl + 2]),
'pol': int(lines[bl + 3]),
'N_ang': int(lines[bl + 4]),
'ang': float(lines[bl + 5]),
'freq': float(lines[bl + 6]),
'FOV': float(lines[bl + 7]),
'FOV_units': s(lines[bl + 8]),
'assim': b(lines[bl + 9]),
'scale': b(lines[bl + 10]),
'getinnov': b(lines[bl + 11]),
'RTM_ID': int(lines[bl + 12]),
'bias_Npar': int(lines[bl + 13]),
'bias_trel': int(lines[bl + 14]),
'bias_tcut': int(lines[bl + 15]),
'nodata': float(lines[bl + 16]),
'varname': s(lines[bl + 17]),
'units': s(lines[bl + 18]),
'path': s(lines[bl + 19]),
'name': s(lines[bl + 20]),
'scalepath': s(lines[bl + 21]),
'scalename': s(lines[bl + 22]),
'flistpath': s(lines[bl + 23]),
'flistname': s(lines[bl + 24]),
'errstd': float(lines[bl + 25]),
'std_normal_max': float(lines[bl + 26]),
'zeromean': b(lines[bl + 27]),
'coarsen_pert': b(lines[bl + 28]),
'xcorr': float(lines[bl + 29]),
'ycorr': float(lines[bl + 30]),
'adapt': int(lines[bl + 31])})
else:
res.append({'descr': s(lines[bl + 0]),
'species': int(lines[bl + 1]),
'orbit': int(lines[bl + 2]),
'pol': int(lines[bl + 3]),
'N_ang': int(lines[bl + 4]),
'ang': float(lines[bl + 5]),
'freq': float(lines[bl + 6]),
'FOV': float(lines[bl + 7]),
'FOV_units': s(lines[bl + 8]),
'assim': b(lines[bl + 9]),
'scale': b(lines[bl + 10]),
'getinnov': b(lines[bl + 11]),
'RTM_ID': int(lines[bl + 12]),
'bias_Npar': int(lines[bl + 13]),
'bias_trel': int(lines[bl + 14]),
'bias_tcut': int(lines[bl + 15]),
'nodata': float(lines[bl + 16]),
'varname': s(lines[bl + 17]),
'units': s(lines[bl + 18]),
'path': s(lines[bl + 19]),
'name': s(lines[bl + 20]),
'scalepath': s(lines[bl + 21]),
'scalename': s(lines[bl + 22]),
'errstd': float(lines[bl + 23]),
'errstd_file': b(lines[bl + 24]),
'path_errstd': s(lines[bl + 25]),
'std_normal_max': float(lines[bl + 26]),
'zeromean': b(lines[bl + 27]),
'coarsen_pert': b(lines[bl + 28]),
'xcorr': float(lines[bl + 29]),
'ycorr': float(lines[bl + 30]),
'adapt': int(lines[bl + 31])})
elif n_fields == 34 and (n_blocks == 4 or n_blocks == 8):
res.append({'descr': s(lines[bl + 0]),
'species': int(lines[bl + 1]),
'orbit': int(lines[bl + 2]),
'pol': int(lines[bl + 3]),
'N_ang': int(lines[bl + 4]),
'ang': float(lines[bl + 5]),
'freq': float(lines[bl + 6]),
'FOV': float(lines[bl + 7]),
'FOV_units': s(lines[bl + 8]),
'assim': b(lines[bl + 9]),
'scale': b(lines[bl + 10]),
'getinnov': b(lines[bl + 11]),
'RTM_ID': int(lines[bl + 12]),
'bias_Npar': int(lines[bl + 13]),
'bias_trel': int(lines[bl + 14]),
'bias_tcut': int(lines[bl + 15]),
'nodata': float(lines[bl + 16]),
'varname': s(lines[bl + 17]),
'units': s(lines[bl + 18]),
'path': s(lines[bl + 19]),
'name': s(lines[bl + 20]),
'scalepath': s(lines[bl + 21]),
'scalename': s(lines[bl + 22]),
'flistpath': s(lines[bl + 23]),
'flistname': s(lines[bl + 24]),
'errstd': float(lines[bl + 25]),
'errstd_file': s(lines[bl + 26]),
'path_errstd': s(lines[bl + 27]),
'std_normal_max': float(lines[bl + 28]),
'zeromean': b(lines[bl + 29]),
'coarsen_pert': b(lines[bl + 30]),
'xcorr': float(lines[bl + 31]),
'ycorr': float(lines[bl + 32]),
'adapt': int(lines[bl + 33])})
else:
res.append({'descr': s(lines[bl + 0]),
'species': int(lines[bl + 1]),
'orbit': int(lines[bl + 2]),
'pol': int(lines[bl + 3]),
'N_ang': int(lines[bl + 4]),
'ang': float(lines[bl + 5]),
'freq': float(lines[bl + 6]),
'FOV': float(lines[bl + 7]),
'FOV_units': s(lines[bl + 8]),
'assim': b(lines[bl + 9]),
'scale': b(lines[bl + 10]),
'getinnov': b(lines[bl + 11]),
'RTM_ID': int(lines[bl + 12]),
'bias_Npar': int(lines[bl + 13]),
'bias_trel': int(lines[bl + 14]),
'bias_tcut': int(lines[bl + 15]),
'nodata': float(lines[bl + 16]),
'varname': s(lines[bl + 17]),
'units': s(lines[bl + 18]),
'path': s(lines[bl + 19]),
'name': s(lines[bl + 20]),
'scalepath': s(lines[bl + 21]),
'scalename': s(lines[bl + 22]),
'errstd': float(lines[bl + 23]),
'std_normal_max': float(lines[bl + 24]),
'zeromean': b(lines[bl + 25]),
'coarsen_pert': b(lines[bl + 26]),
'xcorr': float(lines[bl + 27]),
'ycorr': float(lines[bl + 28]),
'adapt': int(lines[bl + 29])})
return pd.DataFrame(res)
def get_species(self, pol=None, orbit=None, ang=None):
pol = 1 if pol.upper() == 'H' else 2
orbit = 1 if orbit.upper() == 'A' else 2
return self.obsparam.loc[(self.obsparam['pol'] == pol) & \
(self.obsparam['orbit'] == orbit) & \
(self.obsparam['ang'] == ang),'species'].values[0]
def read_fortran_binary(self, fname, dtype,
hdr=None,
length=None,
reg_ftags=True,
loc=None,
idx=None):
"""
Class for reading fortran binary files
Parameters
----------
fname : str
Name of the file to be read
dtype : np.dtype
Template holding parameter names and data types of the file entries
hdr : int
Number of (4-byte) header entries to be skipped
length : int
If provided together with 'hdr':
Position of file length information within the header
else:
Number of successive data blocks contained in the file
reg_ftags : Boolean
If True, a fortran tag (byte) is expected before and after each data field, otherwise
only before and after each data block (only the case for tilegrids files)
loc : int
read only the <loc>-th data element of the file. (only works if reg_ftags is True)
TODO: currently rather useless for ObsFcstAna, because file-lengths don't make much sense
idx : str
If specified, the data of the field named 'idx' will be used as index of the output data frame
Returns
-------
data : pd.DataFrame (if return_hdr is False)
Content of the fortran binary file
"""
if not Path(fname).exists():
logging.warning('file "', fname, '" not found.')
return None
fid = open(fname, 'rb')
# read header
if hdr is not None:
if self.mode == 'GEOSldas':
hdr = np.fromfile(fid, dtype='>i4', count=hdr).newbyteorder()
else:
hdr = np.fromfile(fid, dtype='>i4', count=hdr).byteswap().newbyteorder()
if length is not None:
length = hdr[length]
else:
length = hdr[1]
else:
if length is None:
length = len(self.grid.tilecoord)
if loc is None:
data = pd.DataFrame(columns=dtype.names, index=np.arange(length))
else:
data = pd.DataFrame(columns=dtype.names, index=(loc,))
if length==0:
return data
# read data
if reg_ftags is True:
for dt in dtype.names:
if loc is None:
fid.seek(4, 1) # skip fortran tag
if self.mode == 'GEOSldas':
data.loc[:, dt] = np.fromfile(fid, dtype=dtype[dt], count=length).newbyteorder()
else:
data.loc[:, dt] = np.fromfile(fid, dtype=dtype[dt], count=length).byteswap().newbyteorder()
fid.seek(4, 1) # skip fortran tag
else:
fid.seek(4 + 4*loc, 1)
if self.mode == 'GEOSldas':
data.loc[:, dt] = np.fromfile(fid, dtype=dtype[dt], count=1).newbyteorder()
else:
data.loc[:, dt] = np.fromfile(fid, dtype=dtype[dt], count=1).byteswap().newbyteorder()
fid.seek(4 + 4*length - 4*loc - 4, 1)
else:
for i in np.arange(length):
fid.seek(4, 1) # skip fortran tag
for dt in dtype.names:
if 'S' in str(dtype[dt]):
data.loc[i, dt] = np.fromfile(fid, dtype=dtype[dt], count=1)[0]
else:
if self.mode == 'GEOSldas':
data.loc[i, dt] = np.fromfile(fid, dtype=dtype[dt], count=1)[0].byteswap()
else:
data.loc[i, dt] = np.fromfile(fid, dtype=dtype[dt], count=1)[0]
fid.seek(4, 1) # skip fortran tag
fid.close()
if idx is not None:
data.index = data.loc[:, idx].values
data.drop(idx, axis='columns', inplace=True)
return data
def read_params(self, param, fname=None):
""" Read parameter files (tilegrids, tilecoord, RTMparam, catparam"""
if fname is None:
fname = list(self.paths.rc_out.glob('**/*' + param + '*'))[0]
reg_ftags = False if param == 'tilegrids' else True
dtype, hdr, length = get_template(param)
data = self.read_fortran_binary(fname, dtype, hdr=hdr, length=length, reg_ftags=reg_ftags)
data.replace(-9999., np.nan, inplace=True)
if param == 'tilegrids':
data.index = ['global', 'domain']
else:
# index equals the 'tilenum' which starts at 1!!
data.index += 1
return data
def read_scaling_parameters(self, pentad=1, fname=None, tile_id=None, sensor='SMAP'):
"""
Class for reading scaling files. These hold the observation and model mean and standard deviation, and
number of observations per per pentade.
Parameters
----------
pentad : int
If no tile-id / filename is provided, only one pentad will be read
fname : str
The path to a scaling file that should be read
tile_id : int
If provided, the scaling files for all pentads will be read for the specified tile_id
sensor : str
To distinguish SMAP and SMOS templates
Returns
-------
data : pd.DataFrame
Mean, Std.dev, and N_data for model forecasts and observations.
"""
# TODO: WRONG TREATMENT OF ORBIT DIRECTION! A/D IS IN THE FILENAME BEFORE THE PENTADE!
dtype, hdr, length = get_template('hscale', sensor=sensor)
if tile_id is None:
if fname is not None:
data = self.read_fortran_binary(fname, dtype, hdr=hdr, length=length)
else:
data = self.read_fortran_binary(self.files[pentad], dtype, hdr=hdr, length=length)
# increase index to match tilecoord indices
data.index += 1
else:
pentads = np.arange(73)+1
fields = dtype.names[3:]
data = pd.DataFrame(columns=fields, index=pentads)
for pentad in pentads:
tmp_data = self.read_fortran_binary(self.files[pentad], dtype, hdr=hdr, length=length, idx='tile_id')
data.loc[pentad,fields] = tmp_data.loc[tile_id, fields].values
return data
def read_image(self, yr=None, mo=None, da=None, hr=None, mi=None, species=None, ens_id=None, fname=None):
""""
Read an image for a given date/time(/species/subregion)
If a netCDF file has been created with self.bin2netcdf, the image will be read from this file,
otherwise it is read from the fortran binary file
Parameters
----------
yr, mo, da, hr, mi : int
Date for which the image should be read.
species : int
If provided, only a specific species will be read
No effect when reading fortran binaries!
ens_id : int
If provided, only a specific ensemble member will be read
No effect when reading fortran binaries!
fname: string
Direct input of the filename (instead of inferring from date/time information)
Returns
-------
img : np.ndarray (if netCDF has been created)
Complete 2D or 3D image (with/without species, depending on the specified parameter)
pd.DataFrame (if read from fortran binary)
Only tiles for which data is available
"""
# If netCDF file has been created/loaded, use xarray indexing functions
if hasattr(self, 'images'):
datestr = '%04i-%02i-%02i %02i:%02i' % (yr, mo, da, hr, mi)
tmp_img = self.images
if species is not None:
tmp_img = tmp_img.sel(species=species)
if ens_id is not None:
tmp_img = tmp_img.sel(ens_id=ens_id)
img = tmp_img.sel(time=datestr).values
# Otherwise, read from fortran binary
else:
if fname is None:
if 'monthly' in self.param:
datestr = '%04i%02i' % (yr, mo)
else:
datestr = '%04i%02i%02i_%02i%02i' % (yr, mo, da, hr, mi)
fname = [f for f in self.files if f.name.find(datestr) != -1]
if len(fname) == 0:
logging.warning('No files found for: "' + datestr + '".')
return None
elif len(fname) > 1:
logging.warning('Multiple files found for: "' + datestr + '".')
else:
fname = fname[0]
img = self.read_fortran_binary(fname, self.dtype, hdr=self.hdr, length=self.length)
# set index to match tilecoord indices
if 'obs_tilenum' in img:
# ObsFcstAna files
img.index = img['obs_tilenum'].values
else:
# All other files (hopefully)
img.index += 1
return img
def read_ts(self, param, col, row, species=None, ens_id=None, lonlat=True):
""" Reads a time series from the netCDF time series chunked cube. """
if lonlat is True:
col, row = self.grid.lonlat2colrow(col, row, domain=True)
tmp_ts = self.timeseries[param]
if species is not None:
tmp_ts = tmp_ts.sel(species=species)
if ens_id is not None:
tmp_ts = tmp_ts.sel(ens_id=ens_id)
return tmp_ts.isel(lat=row,lon=col).to_pandas()
@staticmethod
def write_fortran_block(fid, data):
"""" Writes a data block (1D numpy array) into a binary file including fortran tags """
# Force big endian
# if self.mode == 'GEOSldas':
# dtype = data.dtype.newbyteorder('<')
# else:
dtype = data.dtype.newbyteorder('>')
# force 32 bit if 64 bit (usually for float)
if dtype.str[2]=='8':
dtype = np.dtype(dtype.str[0:2]+'4')
ftag = data.size * dtype.itemsize
np.array(ftag).astype('>i4').tofile(fid)
data.astype(dtype).tofile(fid)
np.array(ftag).astype('>i4').tofile(fid)
@staticmethod
def ncfile_init(fname, dimensions, variables):
""""
Method to initialize dimensions/variables of a image-chunked netCDF file
Parameters
----------
fname : str
Filename of the netCDF file to be created
dimensions : dict
Dictionary containing the dimension names and values
variables : list
list of variables to be created with the specified dimensions
Returns
-------
ds : fileid
File ID of the created netCDF file
"""
ds = Dataset(fname, mode='w')
timeunit = 'hours since 2000-01-01 00:00'
# Initialize dimensions
chunksizes = []
for dim in dimensions:
# convert pandas Datetime Index to netCDF-understandable numeric format
if dim == 'time':
dimensions[dim] = date2num(dimensions[dim].to_pydatetime(), timeunit).astype('int32')
# Files are per default image chunked
if dim in ['lon','lat']:
chunksize = len(dimensions[dim])
else:
chunksize = 1
chunksizes.append(chunksize)
dtype = dimensions[dim].dtype
ds.createDimension(dim, len(dimensions[dim]))
ds.createVariable(dim,dtype,
dimensions=(dim,),
chunksizes=(chunksize,),
zlib=True)
ds.variables[dim][:] = dimensions[dim]
# Coordinate attributes following CF-conventions
if 'time' in dimensions:
ds.variables['time'].setncatts({'long_name': 'time',
'units': timeunit})
ds.variables['lon'].setncatts({'long_name': 'longitude',
'units':'degrees_east'})
ds.variables['lat'].setncatts({'long_name': 'latitude',
'units':'degrees_north'})
# Initialize variables
for var in variables:
ds.createVariable(var, 'float32',
dimensions=list(dimensions.keys()),
chunksizes=chunksizes,
fill_value=-9999.,
zlib=True)
return ds
def bin2netcdf(self,
overwrite=False,
date_from=None,
date_to=None,
latmin=-90.,
latmax=90.,
lonmin=-180.,
lonmax=180.,
out_file=None):
""""
Convert fortran binary image into a netCDF data cube.
Parameters
----------
overwrite : boolean
If set, an already existing netCDF file will be overwritten
date_from : string
Lower time limit for which a netCDF image cube should be generated (string format, e.g., '2010-01-01')
date_to : string
Upper time limit for which a netCDF image cube should be generated (string format, e.g., '2010-01-01')
latmin : float
Lower latitude limit for which a netCDF image cube should be generated
latmax : float
Upper latitude limit for which a netCDF image cube should be generated
lonmin : float
Lower longitude limit for which a netCDF image cube should be generated
lonmax : float
Upper longitude limit for which a netCDF image cube should be generated
out_file : string
Optional alternative path / filename for the created NetCDF image cube
"""
if '.nc4' in str(self.files[0]):
logging.warning('bin2netcdf: Wrong method for creating NetCDF cubes. Use "mergenc4files" instead.')
return
if out_file is None:
out_file = self.files[0].parents[3] / (self.param + '_images.nc')
# remove file if it already exists
if hasattr(self,'images'):
if overwrite is False:
logging.warning('bin2netcdf: NetCDF image file already exists. Use keyword "overwrite" to regenerate.')
return
else:
delattr(self, 'images')
out_file.unlink()
# get variable names from fortran reader template
if self.param == 'hscale':
variables = get_template(self.param)[0].names[3::]
else:
variables = get_template(self.param)[0].names
# If specified, only generate netCDF file for specific date range
if self.param == 'hscale':
pentads = np.unique(self.pentads)
orbits = np.unique(self.orbits)
elif self.param == 'ObsFcstAnaEns':
files = self.files
dates = self.dates
ids = self.ens_id
if date_from is not None:
files = files[dates >= pd.to_datetime(date_from)]
ids = ids[dates >= pd.to_datetime(date_from)]
dates = dates[dates >= pd.to_datetime(date_from)]
if date_to is not None:
files = files[dates <= pd.to_datetime(date_to)]
ids = ids[dates <= pd.to_datetime(date_to)]
dates = dates[dates <= pd.to_datetime(date_to)]
else:
dates = self.dates
if date_from is not None:
dates = dates[dates >= pd.to_datetime(date_from)]
if date_to is not None:
dates = dates[dates <= pd.to_datetime(date_to)]
domainlons = self.grid.ease_lons[np.min(self.grid.tilecoord.i_indg):(np.max(self.grid.tilecoord.i_indg)+1)]
domainlats = self.grid.ease_lats[np.min(self.grid.tilecoord.j_indg):(np.max(self.grid.tilecoord.j_indg)+1)]
lonmin = domainlons[np.argmin(np.abs(domainlons-lonmin))]
lonmax = domainlons[np.argmin(np.abs(domainlons-lonmax))]
latmin = domainlats[np.argmin(np.abs(domainlats-latmin))]
latmax = domainlats[np.argmin(np.abs(domainlats-latmax))]
# Use grid lon lat to avoid rounding issues
tmp_tilecoord = self.grid.tilecoord.copy()
tmp_tilecoord['com_lon'] = self.grid.ease_lons[self.grid.tilecoord.i_indg]
tmp_tilecoord['com_lat'] = self.grid.ease_lats[self.grid.tilecoord.j_indg]
# Clip region based on specified coordinate boundaries
ind_img = self.grid.tilecoord[(tmp_tilecoord['com_lon']>=lonmin)&(tmp_tilecoord['com_lon']<=lonmax)&
(tmp_tilecoord['com_lat']<=latmax)&(tmp_tilecoord['com_lat']>=latmin)].index
lons = domainlons[(domainlons >= lonmin) & (domainlons <= lonmax)]
lats = domainlats[(domainlats >= latmin) & (domainlats <= latmax)]
i_offg_2 = np.where(domainlons >= lonmin)[0][0]
j_offg_2 = np.where(domainlats <= latmax)[0][0]
# Innovation file data has an additional 'species' dimension
if self.param == 'ObsFcstAna':
# Remove dates which do not contain any data
nodata = list()
for i, dt in enumerate(dates):
logging.info('%d / %d' % (i, len(dates)))
data = self.read_image(dt.year, dt.month, dt.day, dt.hour, dt.minute)
data = data.loc[data.index.intersection(ind_img), :]
if len(data) == 0:
nodata.append(dt)
dates = dates.drop(nodata)
if len(dates) == 0:
logging.warning('Images do not contain valid data.')
return
spc = pd.DataFrame(self.obsparam)['species'].values.astype('uint8')
dimensions = OrderedDict([('time', dates), ('species', spc), ('lat', lats), ('lon', lons)])
# Innovation ensemble file data has an additional 'species' + 'ens_id' dimension
elif self.param == 'ObsFcstAnaEns':
# Remove dates which do not contain any data
nodata = list()
for i, fn in enumerate(files):
logging.info('%d / %d' % (i, len(files)))
data = self.read_image(fname=fn)
data = data.loc[data.index.intersection(ind_img), :] # clip subregion
if len(data) == 0:
nodata.append(i)
files = np.delete(files, nodata)
dates = np.delete(dates, nodata)
ids = np.delete(ids, nodata)
if len(files) == 0:
logging.warning('Images do not contain valid data.')
return
udates = dates.unique()
uids = np.unique(ids)
spc = pd.DataFrame(self.obsparam)['species'].values.astype('uint8')
dimensions = OrderedDict([('time', udates), ('ens_id', uids), ('species', spc), ('lat', lats), ('lon', lons)])
elif self.param == 'hscale':
dimensions = OrderedDict([('pentad', pentads), ('orbit', orbits), ('lat', lats), ('lon', lons)])
else:
dimensions = OrderedDict([('time', dates), ('lat', lats), ('lon', lons)])
dataset = self.ncfile_init(out_file, dimensions, variables)
if self.param == 'hscale':
for i,(file, pentad, orbit) in enumerate(zip(self.files, self.pentads, self.orbits)):
logging.info('%d / %d' % (i, len(self.files)))
data = self.read_scaling_parameters(fname=file)
data = data.loc[data.index.intersection(ind_img), :]
if len(data) == 0:
continue
img = np.full((1, 1, len(lats), len(lons)), -9999., dtype='float32')
ind_pen = np.where(pentads==pentad)[0][0]
ind_orb = np.where(orbits==orbit)[0][0]
ind_lat = self.grid.tilecoord.loc[ind_img, 'j_indg'].values - self.grid.tilegrids.loc[
'domain', 'j_offg'] - j_offg_2
ind_lon = self.grid.tilecoord.loc[ind_img, 'i_indg'].values - self.grid.tilegrids.loc[
'domain', 'i_offg'] - i_offg_2
for var in variables:
# replace NaN values with the default -9999. fill Value
tmp_img = data[var].values.copy()
np.place(tmp_img, np.isnan(tmp_img), -9999.)
img[0, 0, ind_lat, ind_lon] = tmp_img
dataset.variables[var][ind_pen, ind_orb, :, :] = img
elif self.param == 'ObsFcstAnaEns':
for i,(fn,dt,ensid) in enumerate(zip(files,dates,ids)):
logging.info('%d / %d' % (i, len(files)))
data = self.read_image(fname=fn)
data = data.loc[data.index.intersection(ind_img), :]
if len(data) == 0:
continue
ind_dt = np.where(udates==dt)[0][0]
ind_id = ensid
ind_spc = data['obs_species'].values - 1
ind_lat = self.grid.tilecoord.loc[data['obs_tilenum'].values, 'j_indg'].values - self.grid.tilegrids.loc['domain','j_offg'] - j_offg_2
ind_lon = self.grid.tilecoord.loc[data['obs_tilenum'].values, 'i_indg'].values - self.grid.tilegrids.loc['domain','i_offg'] - i_offg_2
for j,var in enumerate(variables):
# replace NaN values with the default -9999. fill Value
img = np.full((len(spc), len(lats), len(lons)), -9999., dtype='float32')
tmp_img = data[var].values.copy()
np.place(tmp_img, np.isnan(tmp_img), -9999.)
img[ind_spc, ind_lat, ind_lon] = tmp_img
dataset.variables[var][ind_dt, ind_id, :, :, :] = img
else:
for i,dt in enumerate(dates):
logging.info('%d / %d' % (i, len(dates)))
data = self.read_image(dt.year, dt.month, dt.day, dt.hour, dt.minute)
data = data.loc[data.index.intersection(ind_img), :]
if len(data) == 0:
continue
if self.param == 'ObsFcstAna':
img = np.full((len(spc), len(lats), len(lons)), -9999., dtype='float32')
ind_lat = self.grid.tilecoord.loc[data['obs_tilenum'].values, 'j_indg'].values - self.grid.tilegrids.loc['domain','j_offg'] - j_offg_2
ind_lon = self.grid.tilecoord.loc[data['obs_tilenum'].values, 'i_indg'].values - self.grid.tilegrids.loc['domain','i_offg'] - i_offg_2
ind_spc = data['obs_species'].values - 1
else:
img = np.full((len(lats),len(lons)), -9999., dtype='float32')
ind_lat = self.grid.tilecoord.loc[ind_img, 'j_indg'].values - self.grid.tilegrids.loc['domain','j_offg'] - j_offg_2
ind_lon = self.grid.tilecoord.loc[ind_img, 'i_indg'].values - self.grid.tilegrids.loc['domain','i_offg'] - i_offg_2
for var in variables:
# replace NaN values with the default -9999. fill Value
tmp_img = data[var].values.copy()
np.place(tmp_img, np.isnan(tmp_img), -9999.)
if self.param == 'ObsFcstAna':
img[ind_spc,ind_lat,ind_lon] = tmp_img
dataset.variables[var][i,:,:,:] = img
else:
img[ind_lat,ind_lon] = tmp_img
dataset.variables[var][i,:,:] = img
# Save file to disk and loat it as xarray Dataset into the class variable space
dataset.close()
self.images = xr.open_dataset(out_file)
def mergenc4files(self,
overwrite=False,
date_from=None,
date_to=None,
latmin=-90.,
latmax=90.,
lonmin=-180.,
lonmax=180.,
out_file=None):
""""
Convert fortran binary image into a netCDF data cube.
Parameters
----------
overwrite : boolean
If set, an already existing netCDF file will be overwritten
date_from : string
Lower time limit for which a netCDF image cube should be generated (string format, e.g., '2010-01-01')
date_to : string
Upper time limit for which a netCDF image cube should be generated (string format, e.g., '2010-01-01')
latmin : float
Lower latitude limit for which a netCDF image cube should be generated
latmax : float
Upper latitude limit for which a netCDF image cube should be generated
lonmin : float
Lower longitude limit for which a netCDF image cube should be generated
lonmax : float
Upper longitude limit for which a netCDF image cube should be generated
out_file : string
Optional alternative path / filename for the created NetCDF image cube
"""
if out_file is None:
out_file = self.files[0].parents[3] / (self.param + '_images.nc')
# remove file if it already exists
if hasattr(self,'images'):
if overwrite is False:
logging.warning('mergenc4files: NetCDF image file already exists. Use keyword "overwrite" to regenerate.')
return
else:
delattr(self, 'images')
out_file.unlink()
dates = self.dates
files = self.files
if date_from is not None:
files = files[dates >= pd.to_datetime(date_from)]
dates = dates[dates >= pd.to_datetime(date_from)]
if date_to is not None:
files = files[dates <= pd.to_datetime(date_to)]
dates = dates[dates <= pd.to_datetime(date_to)]
domainlons = self.grid.ease_lons[np.min(self.grid.tilecoord.i_indg):(np.max(self.grid.tilecoord.i_indg)+1)]
domainlats = self.grid.ease_lats[np.min(self.grid.tilecoord.j_indg):(np.max(self.grid.tilecoord.j_indg)+1)]
lonmin = domainlons[np.argmin(np.abs(domainlons-lonmin))]
lonmax = domainlons[np.argmin(np.abs(domainlons-lonmax))]
latmin = domainlats[np.argmin(np.abs(domainlats-latmin))]
latmax = domainlats[np.argmin(np.abs(domainlats-latmax))]
# Use grid lon lat to avoid rounding issues
tmp_tilecoord = self.grid.tilecoord.copy()
tmp_tilecoord['com_lon'] = self.grid.ease_lons[self.grid.tilecoord.i_indg]
tmp_tilecoord['com_lat'] = self.grid.ease_lats[self.grid.tilecoord.j_indg]
# Clip region based on specified coordinate boundaries
ind_img = self.grid.tilecoord[(tmp_tilecoord['com_lon']>=lonmin)&(tmp_tilecoord['com_lon']<=lonmax)&
(tmp_tilecoord['com_lat']<=latmax)&(tmp_tilecoord['com_lat']>=latmin)].index
lons = domainlons[(domainlons >= lonmin) & (domainlons <= lonmax)]
lats = domainlats[(domainlats >= latmin) & (domainlats <= latmax)]
i_offg_2 = np.where(domainlons >= lonmin)[0][0]
j_offg_2 = np.where(domainlats <= latmax)[0][0]
dimensions = OrderedDict([('time', dates), ('lat', lats), ('lon', lons)])
with Dataset(self.files[0]) as ds:
variables = set(ds.variables.keys()).difference(['JG','IG','lon','lat'])