-
Notifications
You must be signed in to change notification settings - Fork 0
/
read_file.py
1396 lines (1222 loc) · 77.7 KB
/
read_file.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
"""
@author: debashri
This file contains different utility fuctions to help the anomaly_detection_main.py
write_pickle: writes the radar samples in the pickle file
load_nonconjugate_feat: loads non-conjugate features
read_nonconjugate_feat: reads non-conjugate features
load_conjugate_feat: loads conjugate features
read_conjugate_feat: reads conjugate features
"""
import pandas as pd
from sklearn import preprocessing
import pickle
from random import choices
import os
import re
import numpy as np
import scipy
import pandas as pd
import glob
#from random import random
import random
from collections import defaultdict
from pathlib import Path
def read_and_extract_metadata_NEU_dataset(metadata_path, meta_filename, feature_shape, num_classes):
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1,
how='all')
one_file_label = np.zeros((feature_shape, num_classes)) # LTE, DSSS-BPSK, DSSS-QPSK, DSSS spreading gain (5) - total 8
if 'LTE_DSSS' in meta_filename:
# print("one_label_file_df: ", one_label_file_df)
# print("Columns: ", list(one_label_file_df.columns))
# print("DSSS Modulation: ", one_label_file_df['DSSS_Mod'])
# print("DSSS Spreading Gain: ", one_label_file_df['DSSS_PolyDegree_M'])
one_file_label[:, 0] = 1
# if one_label_file_df['DSSS_Mod'].to_numpy() == 'BPSK':
# one_file_label[:, 1] = 1
if one_label_file_df['DSSS_Mod'].to_numpy() == 'QPSK':
one_file_label[:, 1] = 1
# considering the spreading gains
if one_label_file_df['DSSS_SpreadingGain'].to_numpy() == 63:
one_file_label[:, 2] = 1
if one_label_file_df['DSSS_SpreadingGain'].to_numpy() == 127:
one_file_label[:, 3] = 1
if one_label_file_df['DSSS_SpreadingGain'].to_numpy() == 255:
one_file_label[:, 4] = 1
if one_label_file_df['DSSS_SpreadingGain'].to_numpy() == 511:
one_file_label[:, 5] = 1
if one_label_file_df['DSSS_SpreadingGain'].to_numpy() == 1023:
one_file_label[:, 6] = 1
# SIR tag: 'LTE_DSSS_SIR_dB'
# lte_dsss_count = lte_dsss_count + 1
# else: # Only LTE
# pass
# one_file_label[:, 0] = 1
# only_lte_count = only_lte_count + 1
# print("For ", meta_filename, " the labels are: ", one_file_label)
return one_file_label
def extract_snr_from_metadata_NEU_dataset(metadata_path, meta_filename):
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1,
how='all')
return one_label_file_df['LTE_SNR_dB'].to_numpy()
def extract_sir_from_metadata_NEU_dataset(metadata_path, meta_filename):
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1,
how='all')
return one_label_file_df['LTE_DSSS_SIR_dB'].to_numpy()
# Generate the labels from meta data
def generate_labels_NEU_dataset(metadata_path, meta_filename, num_classes):
one_file_label = np.zeros((1, num_classes))
# print("The metadata file name: ", (metadata_path + meta_filename + '.csv'))
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1, how='all')
if one_label_file_df['Signal_Type'].to_numpy() == 'LTE':
one_file_label[0, 0] = 1
if one_label_file_df['Signal_Type'].to_numpy() == 'LTE_DSSS':
one_file_label[0, 1] = 1
return one_file_label
# extract the statistics about the CSP features
def extract_statistic_from_CSP_features(CSP_features_one_file):
mean = np.max(CSP_features_one_file, axis=0)
return mean
# extract the statistics about the CSP features
def extract_statistic_from_CSP_nc_features_one_col(CSP_features_one_file):
# print("coming here")
row_with_max = np.argmax(CSP_features_one_file[:,1], axis=0) # for non-conjugate features 2nd column, for conjugate features 4th column is important
# print("important information..", row_with_max, CSP_features_one_file[row_with_max, :], CSP_features_one_file)
return CSP_features_one_file[row_with_max, :]
# extract the statistics about the CSP features
def extract_statistic_from_CSP_c_features_one_col(CSP_features_one_file):
# print("coming here")
row_with_max = np.argmax(CSP_features_one_file[:,3], axis=0) # for non-conjugate features 2nd column, for conjugate features 4th column is important
# print("important information..", row_with_max, CSP_features_one_file[row_with_max, :], CSP_features_one_file)
return CSP_features_one_file[row_with_max, :]
# generate one CSP feature in form of 2D matrix
def organize_CSP_features_2D(CSP_features_one_file):
channels = CSP_features_one_file.shape[1]
rows_to_be_taken = 100
rows_to_be_added = 0
columnIndex = 1 # sorting by the second column
CSP_features_one_file = CSP_features_one_file[(-CSP_features_one_file[:, columnIndex]).argsort()] # sort the array in decending order by second column
if CSP_features_one_file.shape[0] <rows_to_be_taken:
rows_to_be_added = rows_to_be_taken - CSP_features_one_file.shape[0]
else:
CSP_features_one_file = CSP_features_one_file[0:rows_to_be_taken, :]
# print("shape 1: ", CSP_features_one_file.shape)
CSP_features_one_file = np.pad(CSP_features_one_file, [(0, rows_to_be_added), (0, 0)], mode='constant', constant_values=0)
# print("******testing*******: ", CSP_features_one_file.shape)
return CSP_features_one_file
# generate one CSP feature in form of 3D matrix
def organize_CSP_features_3D(CSP_features_one_file):
channels = CSP_features_one_file.shape[1]
rows_to_be_taken = 100
rows_to_be_added = 0
columnIndex = 1 # sorting by the second column
# CSP_features_one_file = CSP_features_one_file[np.argsort(CSP_features_one_file)] #[-rows_to_be_taken:]
# CSP_features_one_file = CSP_features_one_file[CSP_features_one_file[:, columnIndex].argsort()]
CSP_features_one_file = CSP_features_one_file[(-CSP_features_one_file[:, columnIndex]).argsort()] # sort the array in decending order by second column
if CSP_features_one_file.shape[0] <rows_to_be_taken:
rows_to_be_added = rows_to_be_taken - CSP_features_one_file.shape[0]
else:
CSP_features_one_file = CSP_features_one_file[0:rows_to_be_taken, :]
# print("shape 1: ", CSP_features_one_file.shape)
CSP_features_one_file = np.pad(CSP_features_one_file, [(0, rows_to_be_added), (0, 0)], mode='constant', constant_values=0)
# print("shape 2: ", CSP_features_one_file.shape)
columns = CSP_features_one_file.shape[0]
CSP_features_one_file = np.expand_dims(CSP_features_one_file, axis=1)
# print("shape 3: ", CSP_features_one_file.shape)
CSP_features_one_file = np.swapaxes (CSP_features_one_file, 0, 1)
# print("shape 4: ", CSP_features_one_file.shape)
# CSP_features_one_file = np.swapaxes(CSP_features_one_file, 0, 1)
# print("shape 5: ", CSP_features_one_file.shape)
# CSP_features_one_file = np.reshape(CSP_features_one_file, (1, columns, channels))
return CSP_features_one_file
# read one IQ file
def read_one_iq(iq_path, filename, block_length, slicing, slice_length):
dtype_all = scipy.dtype([('raw-iq', scipy.complex64)])
with open(iq_path + filename, mode='rb') as file: # b is important -> binary
iqdata_one_file = scipy.fromfile(file, dtype=dtype_all)
# if slicing == True:
# slice_index = random.randint(0, iqdata_one_file.shape[0] - slice_length - 1)
# iqdata_one_file = iqdata_one_file[slice_index:slice_index + slice_length, :]
#
# # slice_index = random.randint(0, iqdata_one_file.shape[0] - slice_length - 1)
# # iqdata_one_file = iqdata_one_file[slice_index:slice_index + slice_length]
# # iqdata_one_file = np.expand_dims(iqdata_one_file, axis=1)
# # iqdata_one_file = np.transpose(iqdata_one_file)
# else: # reading the whole i/q file stacked by different block lengths (block_len)
iqdata_one_file = np.reshape(iqdata_one_file[:(iqdata_one_file.shape[0] // block_length) * block_length], (iqdata_one_file.shape[0] // block_length, block_length)) # discard the extra elements for uneven array
# print("Shape before: ", iqdata_one_file.shape)
if slicing == True:
slice_index = random.randint(0, iqdata_one_file.shape[1] - slice_length - 1)
iqdata_one_file = iqdata_one_file[:, slice_index:slice_index + slice_length]
# print("After: ", iqdata_one_file.shape)
iqdata_one_file = np.expand_dims(iqdata_one_file, axis=2)
iqdata_one_file = np.concatenate([iqdata_one_file['raw-iq'].real,
iqdata_one_file['raw-iq'].imag], axis=2)
return iqdata_one_file
#read one CSP (either conjugate/non-conjugate) feature file
def read_one_CSP_feature(filename, feature_options):
try:
one_file_df = pd.read_csv(filename, sep=" ", header=None).dropna(axis=1, how='all')
column_index = list(one_file_df.columns)
selected_columns = [column_index[i] for i in feature_options]
# print("one_df and input shape:", one_file_df[selected_columns].to_numpy().shape, inputs.shape)
one_file_np = one_file_df[selected_columns].to_numpy()
except pd.errors.EmptyDataError:
print("Note:", filename, "****************was empty. Skipping.")
# count = count - 1 # decreasing the total file count
# one_file_df =
one_file_np = np.zeros((1, len(feature_options)))
return one_file_np
def extract_metadata_dsss_variant_from_POWDER_dataset(metadata_path, meta_filename):
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1,
how='all')
one_file_label = np.zeros((1, 2)) # DSSS-BPSK, DSSS-QPSK
if 'LTE_DSSS' in meta_filename:
if one_label_file_df['DSSS_Mod'].to_numpy() == 2:
one_file_label[:, 0] = 1
elif one_label_file_df['DSSS_Mod'].to_numpy() == 4:
one_file_label[:, 1] = 1
return one_file_label
def extract_metadata_from_NWRA_dataset(filename, dsss_type):
loop_index = int(filename.split('_')[6]) # finding the loop index in the data
# print("FLAG: ", filename, flag)
# print("The loop index and dsss_type: ", loop_index, dsss_type)
if loop_index%2 == 0: # loop index is even - only LTE
# print("returning the loop index for lte: ", loop_index)
return 'LTE'
else: # loop index is odd - DSSS present
if dsss_type == 'synthetic' and loop_index > 40:
# print("returning the loop index for dsss: ", loop_index)
return 'SKIP'
if dsss_type == 'real' and loop_index < 40:
return 'SKIP'
# print("returning the loop index for dsss: ", loop_index)
return 'LTE_DSSS'
def extract_metadata_dsss_variant_from_NEU_dataset(metadata_path, meta_filename):
one_label_file_df = pd.read_csv(metadata_path + meta_filename + '.csv', sep=",", header=0).dropna(axis=1,
how='all')
one_file_label = np.zeros((1, 2)) # DSSS-BPSK, DSSS-QPSK
if 'LTE_DSSS' in meta_filename:
if one_label_file_df['DSSS_Mod'].to_numpy() == 'BPSK':
one_file_label[:, 0] = 1
else: one_file_label[:, 1] = 1
return one_file_label
def extract_metadata_dsss_variant_from_NWRA_dataset(foldername, filename, no_of_blocks):
synthetic_dsss_metadata_file_name = foldername+ 'dsss_signal_params.txt'
captured_dsss_metadata_file_name = foldername + 'dsss_signal_params_2.txt'
col1 = 'loop_index'
col2 = 'signal_index'
col3 = 'M_value'
col4 = 'chip_rate'
col5 = 'chipping_seq_index'
col6 = 'CF_offset'
col7 = 'EBW'
col8 = 'power_slicing_factor'
col9 = 'SIR_db'
col10 = 'delay'
col11 = 'upsample'
col12 = 'downsample'
col13 = 'noise_spectral_density'
col14 = 'mod_type' # 1 == BPSK, 2 == QPSK, 3 == SQPSK.
synthetic_dsss_metadata_df = pd.read_csv(synthetic_dsss_metadata_file_name, sep="\s+", header=None,
names=[col1, col2, col3, col4, col5, col6, col7, col8, col9, col10, col11, col12, col13, col14])
captured_dsss_metadata_df = pd.read_csv(captured_dsss_metadata_file_name, sep="\s+", header=None,
names=[col1, col2, 'start_sample', col6, col8, col9, col10, col11, col12, col13])
one_file_label = np.zeros((no_of_blocks, 2))
loop_index = int(filename.split('_')[6])
signal_index = int(filename.split('_')[7].split('.')[0])
#print("*******LOOP INDEX AND SIGNAL INDEX: ***********: ", loop_index, signal_index)
if loop_index % 2 != 0: # DSSS signal, now extract metadata information
if loop_index > 40: # real
#for index, row in captured_dsss_metadata_df.iterrows():
#print("The row: ", row)
#if loop_index == row['loop_index'] and signal_index == row['signal_index']:
one_file_label[:, 0] = 1 # all real DSSS are BPSK only
else: # synthetic
for index, row in synthetic_dsss_metadata_df.iterrows():
#print("The row: ", row)
#print("testing in the function for :", filename, loop_index, row['loop_index'], signal_index, row['signal_index'])
if loop_index == int(row['loop_index']) and signal_index == int(row['signal_index']):
#print("*******************************Entered****************:", loop_index, row['loop_index'], signal_index, row['signal_index'])
if int(row['mod_type']) == 1:
one_file_label[:, 0] = 1
else:
one_file_label[:, 1] = 1 # two options: QPSK and SQPSK
return one_file_label
# def extract_CSP_metadata_from_NWRA_dataset(filename, dsss_type):
# loop_index = int(filename.split('_')[6]) # finding the loop index in the data
# # print("FLAG: ", filename, flag)
# # print("The loop index and dsss_type: ", loop_index, dsss_type)
# if loop_index%2 == 0: # loop index is even - only LTE
# # print("returning the loop index for lte: ", loop_index)
# return 'LTE'
# else: # loop index is odd - DSSS present
# if dsss_type == 'synthetic' and loop_index > 40:
# # print("returning the loop index for dsss: ", loop_index)
# return 'SKIP'
# if dsss_type == 'real' and loop_index < 40:
# return 'SKIP'
# # print("returning the loop index for dsss: ", loop_index)
# return 'LTE_DSSS'
#
# reading all the folders
def generate_inputs_labels_IQ_NC_C_POWDER_dataset(feature_path, input_options = ['iq', 'nc'], feature_options=[0], block_length = 256, num_classes=2, strategy = 4, percentage_to_read = 0.1, slicing = True, slice_length = 256):
metadata_path_5mhz = feature_path + 'POWDER_data/Batch1_5MHz/Metadata/' # 5MHz BW data
iq_path_5mhz = feature_path + 'POWDER_data/Batch1_5MHz/IQ/' # 5MHz BW data
metadata_path_10mhz = feature_path + 'POWDER_data/Batch1_10MHz/Metadata/' # 5MHz BW data
iq_path_10mhz = feature_path + 'POWDER_data/Batch1_10MHz/IQ/' # 5MHz BW data
CSP_path_5mhz = feature_path + 'POWDER_data/Batch1_5MHz/'+str(block_length)+'/'
CSP_path_10mhz = feature_path + 'POWDER_data/Batch1_10MHz/'+str(block_length)+'/'
if block_length < 460800: iq_data_5mhz, csp_c_features_5mhz, csp_nc_features_5mhz, labels_5mhz = read_one_bandwidth_POWDER_dataset(iq_path_5mhz, metadata_path_5mhz, CSP_path_5mhz, input_options, feature_options, block_length,
num_classes, strategy, percentage_to_read/2, slicing, slice_length)
iq_data_10mhz, csp_c_features_10mhz, csp_nc_features_10mhz, labels_10mhz = read_one_bandwidth_POWDER_dataset(iq_path_10mhz, metadata_path_10mhz, CSP_path_10mhz, input_options, feature_options, block_length,
num_classes, strategy, percentage_to_read/2, slicing, slice_length)
if block_length < 460800:
iq_data = np.concatenate((iq_data_5mhz, iq_data_10mhz), axis=0)
csp_c_features= np.concatenate((csp_c_features_5mhz, csp_c_features_10mhz), axis=0)
csp_nc_features= np.concatenate((csp_nc_features_5mhz, csp_nc_features_10mhz), axis=0)
labels= np.concatenate((labels_5mhz, labels_10mhz), axis=0)
else:
iq_data = iq_data_10mhz
csp_c_features= csp_c_features_10mhz
csp_nc_features= csp_nc_features_10mhz
labels= labels_10mhz
return iq_data, csp_c_features, csp_nc_features, labels
#genearte the input and labels for all the modalities
def read_one_bandwidth_POWDER_dataset(iq_path,metadata_path, CSP_path, input_options = ['iq', 'nc'], feature_options=[0], block_length = 256, num_classes=2, strategy = 4, percentage_to_read = 0.1, slicing = True, slice_length = 256):
iq_data = np.zeros(1)
csp_c_features = np.zeros(1)
csp_nc_features = np.zeros(1)
labels = np.zeros(1)
count_files = 0
for filename in os.listdir(iq_path):
if random.random() > percentage_to_read:
# print("Continuing, ", filename)
continue
meta_filename = filename[::-1].split('_', 0)[-1][::-1]
#print("Meta_filename: ", meta_filename)
#if 'iq' in input_options:
iqdata_one_file = read_one_iq(iq_path, filename, block_length, slicing, slice_length)
print("Length of the iq for ", filename, " is:", iqdata_one_file.shape)
# if 'c' in input_options:
extrated_stats_c = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
for file_name in glob.iglob(CSP_path+filename+'_'+str(block_length)+'*.C', recursive = True):
# print("The conjugate feature file name: ", file_name)
one_c_csp_file_np = read_one_CSP_feature(file_name, feature_options)
if strategy == 2:
one_c_csp_file_np = extract_statistic_from_CSP_features(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
elif strategy == 4:
one_c_csp_file_np = extract_statistic_from_CSP_c_features_one_col(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
else:
print(
"Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
extrated_stats_c[csp_feature_count,:] = one_c_csp_file_np
csp_feature_count +=1
print("Shape of conjugate CSP features for ", file_name, ":", one_c_csp_file_np.shape, one_c_csp_file_np.shape, extrated_stats_c.shape)
#if 'nc' in input_options:
extrated_stats_nc = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
for file_name in glob.iglob(CSP_path+filename+'_'+str(block_length)+'*.NC', recursive = True):
# print("The non conjugate feature file name: ", file_name)
one_nc_csp_file_np = read_one_CSP_feature(file_name, feature_options)
################## (without any strategy implementation (old implementation) #######################
# extrated_stats_one_file_nc = extract_statistic_from_CSP_features(one_nc_csp_file_np)
################## (end of without any strategy implementation (old implementation) #######################
########################### new implementation #########################
if strategy == 2:
one_nc_csp_file_np = extract_statistic_from_CSP_features(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
elif strategy == 4:
one_nc_csp_file_np = extract_statistic_from_CSP_nc_features_one_col(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
else:
print("Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
extrated_stats_nc[csp_feature_count, :] = one_nc_csp_file_np
csp_feature_count += 1
print("Shape of conjugate CSP features for ", file_name, ":", one_nc_csp_file_np.shape,
one_nc_csp_file_np.shape, extrated_stats_nc.shape)
# read the corresponding label
one_label = generate_labels_NEU_dataset(metadata_path, meta_filename, num_classes)
one_label = np.repeat(one_label, iqdata_one_file.shape[0], axis=0)
#print("To prepare for Skipping this DSSS file****************** ", filename, one_label[0, 1])
if one_label[0, 1] == 1 and random.random() > 0.25: #to randomly select only one of the DSSS frame for balanced training
print("*******************************Skipping this DSSS file****************** ", filename)
continue
if count_files == 0:
iq_data = iqdata_one_file
csp_c_features = extrated_stats_c
csp_nc_features = extrated_stats_nc
labels = one_label
else:
iq_data = np.concatenate((iq_data, iqdata_one_file), axis=0)
csp_c_features = np.concatenate((csp_c_features, extrated_stats_c), axis=0)
csp_nc_features = np.concatenate((csp_nc_features, extrated_stats_nc), axis=0)
labels = np.concatenate((labels, one_label), axis=0)
# print("the generated label for ", filename, " is:", one_label)
# print("Length of the iq and CSP features for ", filename, " is:", iqdata_one_file.shape, one_c_csp_file_np.shape, one_nc_csp_file_np.shape, one_label.shape)
count_files +=1
return iq_data, csp_c_features, csp_nc_features, labels
#genearte the input and labels for all the modalities
def generate_inputs_labels_IQ_NC_C_NEU_dataset(feature_path, input_options = ['iq', 'nc', 'c'], feature_options=[0], block_length = 256, num_classes=2, snr_list = [0, 5, 10], sir_list=[0, 5, 10], strategy = 4, percentage_to_read = 0.1, slicing = True, slice_length = 2048):
iq_data = np.zeros(1)
csp_c_features = np.zeros(1)
csp_nc_features = np.zeros(1)
labels = np.zeros(1)
metadata_path = feature_path + 'IQDataSet_LTE_DSSS_v2/Metadata/'
# ORIGINAL PATH
iq_path = feature_path + 'IQDataSet_LTE_DSSS_v2/IQ/'
CSP_path = feature_path + 'NEU_LTE_DSSS_Dataset_2_CSP/'
# TEST PATH
# iq_path = feature_path + 'test/IQ/'
# CSP_path = feature_path + 'test/CSP/'
count_files = 0
for filename in os.listdir(iq_path):
if random.random() > percentage_to_read:
# print("Continuing, ", filename)
continue
sir_for_this_file = sir_list[0] # intitlizing the SIR values for only LTE signals
meta_filename = filename[::-1].split('_', 0)[-1][::-1]
# print("Meta_filename: ", meta_filename)
snr_for_this_file = extract_snr_from_metadata_NEU_dataset(metadata_path, meta_filename)
if 'LTE_DSSS' in filename:
sir_for_this_file = extract_sir_from_metadata_NEU_dataset(metadata_path, meta_filename)
# print("THE SIR VALUES: ", sir_for_this_file, sir_list)
# print("**********Before Entering: ", filename, sir_for_this_file, sir_list, snr_for_this_file, snr_list)
if ((snr_for_this_file in snr_list) and (sir_for_this_file in sir_list)):
# read the iq file
if 'iq' in input_options:
iqdata_one_file = read_one_iq(iq_path, filename, block_length, slicing, slice_length)
print("Length of the iq for ", filename, " is:", iqdata_one_file.shape)
# adding to the output array
if count_files == 0: iq_data = iqdata_one_file
else: iq_data = np.concatenate((iq_data, iqdata_one_file), axis=0)
if 'c' in input_options:
extrated_stats_c = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
for file_name in glob.iglob(CSP_path+filename+'_'+str(block_length)+'*.C', recursive = True):
# print("The conjugate feature file name: ", file_name)
one_c_csp_file_np = read_one_CSP_feature(file_name, feature_options)
################## (without any strategy implementation (old implementation) #######################
# extrated_stats_one_file_c = extract_statistic_from_CSP_features(one_c_csp_file_np)
################## (end of without any strategy implementation (old implementation) #######################
########################### new implementation #########################
if strategy == 2:
one_c_csp_file_np = extract_statistic_from_CSP_features(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
elif strategy == 4:
one_c_csp_file_np = extract_statistic_from_CSP_c_features_one_col(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
else:
print(
"Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
extrated_stats_c[csp_feature_count,:] = one_c_csp_file_np
csp_feature_count +=1
print("Shape of conjugate CSP features for ", file_name, ":", one_c_csp_file_np.shape, one_c_csp_file_np.shape, extrated_stats_c.shape)
# adding to the cycle features
if count_files == 0:
csp_c_features = extrated_stats_c
else:
csp_c_features = np.concatenate((csp_c_features, extrated_stats_c), axis=0)
if 'nc' in input_options:
extrated_stats_nc = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
for file_name in glob.iglob(CSP_path+filename+'_'+str(block_length)+'*.NC', recursive = True):
# print("The non conjugate feature file name: ", file_name)
one_nc_csp_file_np = read_one_CSP_feature(file_name, feature_options)
################## (without any strategy implementation (old implementation) #######################
# extrated_stats_one_file_nc = extract_statistic_from_CSP_features(one_nc_csp_file_np)
################## (end of without any strategy implementation (old implementation) #######################
########################### new implementation #########################
if strategy == 2:
one_nc_csp_file_np = extract_statistic_from_CSP_features(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
elif strategy == 4:
one_nc_csp_file_np = extract_statistic_from_CSP_nc_features_one_col(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
else:
print("Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
extrated_stats_nc[csp_feature_count, :] = one_nc_csp_file_np
csp_feature_count += 1
print("Shape of conjugate CSP features for ", file_name, ":", one_nc_csp_file_np.shape,
one_nc_csp_file_np.shape, extrated_stats_nc.shape)
# adding to the cycle features
if count_files == 0:
csp_nc_features = extrated_stats_nc
else:
csp_nc_features = np.concatenate((csp_nc_features, extrated_stats_nc), axis=0)
# read the corresponding label
one_label = generate_labels_NEU_dataset(metadata_path, meta_filename, num_classes)
one_label = np.repeat(one_label, iqdata_one_file.shape[0], axis=0)
if count_files == 0:
labels = one_label
else:
labels = np.concatenate((labels, one_label), axis=0)
# print("the generated label for ", filename, " is:", one_label)
# print("Length of the iq and CSP features for ", filename, " is:", iqdata_one_file.shape, one_c_csp_file_np.shape, one_nc_csp_file_np.shape, one_label.shape)
count_files +=1
return iq_data, csp_c_features, csp_nc_features, labels
def generate_inputs_labels_IQ_NC_C_for_NWRA_dataset(feature_path, input_options = ['iq', 'nc', 'c'], feature_options=[0], block_length = 256, num_classes=2, snr_list = [0, 5, 10], sir_list=[0, 5, 10], strategy = 4, dsss_type ='real', percentage_to_read = 0.1, slicing = True, slice_length = 2048):
iq_data = np.zeros(1)
csp_c_features = np.zeros(1)
csp_nc_features = np.zeros(1)
labels = np.zeros(1)
# metadata_path = feature_path + 'IQDataSet_LTE_DSSS_v2/Metadata/'
# ORIGINAL PATH
iq_path = feature_path + 'NWRA_data/IQ/'
CSP_path = feature_path + 'NWRA_data/'+str(block_length)+'/'
# TEST PATH
# iq_path = feature_path + 'test/IQ/'
# CSP_path = feature_path + 'test/CSP/'
count_files = 0
only_lte_count = 0
lte_dsss_count = 0
for filename in os.listdir(iq_path):
if len(filename.split('_')) <= 6:
continue
if random.random() > percentage_to_read:
# print("Continuing, ", filename)
continue
if 'iq' in input_options:
# with open(iq_path + filename, mode='rb') as file: # b is important -> binary
iqdata_one_file = read_one_iq(iq_path, filename, block_length, slicing, slice_length)
# iqdata_one_file = iqdata_one_file[2:] # skipping first two elements (as per Chad's binary.m file)
# iqdata_one_file = read_one_iq(iq_path, filename, slice_len)
print("Length of the iq for ", filename, " is:", iqdata_one_file.shape)
# adding to the output array
# if count_files == 0: iq_data = iqdata_one_file
# else: iq_data = np.concatenate((iq_data, iqdata_one_file), axis=0)
# read the corresponding label
# Creating the labels: 2 Label version
one_file_label = np.zeros((iqdata_one_file.shape[0], num_classes))
if num_classes == 2:
flag = extract_metadata_from_NWRA_dataset(filename, dsss_type)
if 'SKIP' in flag:
continue
elif 'LTE_DSSS' in flag:
one_file_label[:, 1] = 1
lte_dsss_count = lte_dsss_count + 1
else:
one_file_label[:, 0] = 1
only_lte_count = only_lte_count + 1
# adding /Q data and labels
if count_files == 0:
iq_data = iqdata_one_file
labels = one_file_label
# print("coming here too..", labels.shape)
else:
iq_data = np.concatenate((iq_data, iqdata_one_file), axis=0)
labels = np.concatenate((labels, one_file_label), axis=0)
if 'c' in input_options:
extrated_stats_c = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
for file_name in glob.iglob(CSP_path+os.path.splitext(filename)[0]+'_'+str(block_length)+'*.C', recursive = True):
# print("The conjugate feature file name: ", file_name)
one_c_csp_file_np = read_one_CSP_feature(file_name, feature_options)
################## (without any strategy implementation (old implementation) #######################
# extrated_stats_one_file_c = extract_statistic_from_CSP_features(one_c_csp_file_np)
################## (end of without any strategy implementation (old implementation) #######################
########################### new implementation #########################
if strategy == 2:
one_c_csp_file_np = extract_statistic_from_CSP_features(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
elif strategy == 4:
one_c_csp_file_np = extract_statistic_from_CSP_nc_features_one_col(one_c_csp_file_np)
one_c_csp_file_np = np.expand_dims(one_c_csp_file_np, axis=0)
else:
print("Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
extrated_stats_c[csp_feature_count,:] = one_c_csp_file_np
csp_feature_count +=1
print("Shape of conjugate CSP features for ", file_name, ":", one_c_csp_file_np.shape, one_c_csp_file_np.shape, extrated_stats_c.shape)
# adding to the cycle features
if count_files == 0:
csp_c_features = extrated_stats_c
else:
csp_c_features = np.concatenate((csp_c_features, extrated_stats_c), axis=0)
if 'nc' in input_options:
extrated_stats_nc = np.zeros((iqdata_one_file.shape[0], len(feature_options)))
csp_feature_count = 0
# print("TEST: ", os.path.splitext(filename)[0])
for file_name in glob.iglob(CSP_path+os.path.splitext(filename)[0]+'_'+str(block_length)+'*.NC', recursive = True):
# print("The non conjugate feature file name: ", file_name)
one_nc_csp_file_np = read_one_CSP_feature(file_name, feature_options)
################## (without any strategy implementation (old implementation) #######################
# extrated_stats_one_file_nc = extract_statistic_from_CSP_features(one_nc_csp_file_np)
################## (end of without any strategy implementation (old implementation) #######################
########################### new implementation #########################
if strategy == 2:
one_nc_csp_file_np = extract_statistic_from_CSP_features(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
elif strategy == 4:
one_nc_csp_file_np = extract_statistic_from_CSP_nc_features_one_col(one_nc_csp_file_np)
one_nc_csp_file_np = np.expand_dims(one_nc_csp_file_np, axis=0)
else:
print("Invalid strategy for handling CSP Features while using fusion, please use either 2 or 4.")
exit(0)
########################### end of new implementation #########################
# print("coming here")
#print("Testing the shape: ", one_nc_csp_file_np.shape, one_nc_csp_file_np.squeeze().shape, extrated_stats_nc.shape, csp_feature_count)
extrated_stats_nc[csp_feature_count, :] = one_nc_csp_file_np.squeeze()
csp_feature_count += 1
print("Shape of conjugate CSP features for ", file_name, ":", one_nc_csp_file_np.shape,
one_nc_csp_file_np.shape, extrated_stats_nc.shape)
# adding to the cycle features
if count_files == 0:
csp_nc_features = extrated_stats_nc
else:
csp_nc_features = np.concatenate((csp_nc_features, extrated_stats_nc), axis=0)
# print("the generated label for ", filename, " is:", one_label)
# print("Length of the iq and CSP features for ", filename, " is:", iqdata_one_file.shape, one_c_csp_file_np.shape, one_nc_csp_file_np.shape, one_label.shape)
count_files +=1
return iq_data, csp_c_features, csp_nc_features, labels
def generate_inputs_labels_IQ_for_NEU_dataset(feature_path, blocks = [131072], block_length = 256, num_classes=2, snr_list = [0, 5, 10], sir_list=[0, 5, 10], percentage_to_read = 0.1, slicing = True, slice_length = 2048):
inputs = np.zeros(1)
labels = np.zeros(1)
metadata_path = feature_path + 'IQDataSet_LTE_DSSS_v2/Metadata/'
iq_path = feature_path + 'IQDataSet_LTE_DSSS_v2/IQ/'
# iq_path = feature_path + 'test/IQ/'
count = 0
only_lte_count = 0
lte_dsss_count = 0
dtype_all = scipy.dtype([('raw-iq', scipy.complex64)]) # gr_complex is '32fc' --> make any sense?
for filename in os.listdir(iq_path):
if random.random() > percentage_to_read:
# print("Continuing, ", filename)
continue
sir_for_this_file = sir_list[0] # intitlizing the SIR values for only LTE signals
meta_filename = filename[::-1].split('_', 0)[-1][::-1]
snr_for_this_file = extract_snr_from_metadata_NEU_dataset(metadata_path, meta_filename)
if 'LTE_DSSS' in filename:
sir_for_this_file = extract_sir_from_metadata_NEU_dataset(metadata_path, meta_filename)
# print("THE SIR VALUES: ", sir_for_this_file, sir_list)
# print("**********Before Entering: ", filename, sir_for_this_file, sir_list, snr_for_this_file, snr_list)
if ((snr_for_this_file in snr_list) and (sir_for_this_file in sir_list)): # the second part of the condition is not being used for 'only lte' signals
# print("**********Entering: ", filename, sir_for_this_file, sir_list, snr_for_this_file, snr_list)
with open(iq_path+filename, mode='rb') as file: # b is important -> binary
iqdata_one_file = scipy.fromfile(file, dtype=dtype_all)
if slicing == True:
# slice_index = random.randint(0, iqdata_one_file.shape[0] - slice_length - 1)
# iqdata_one_file = iqdata_one_file[slice_index:slice_index + slice_length, :]
slice_index = random.randint(0, iqdata_one_file.shape[0] - slice_length - 1)
iqdata_one_file = iqdata_one_file[slice_index:slice_index + slice_length]
iqdata_one_file = np.expand_dims(iqdata_one_file, axis=1)
iqdata_one_file = np.transpose(iqdata_one_file)
else: # reading the whole i/q file stacked by different block lengths (block_len)
iqdata_one_file = np.reshape(iqdata_one_file[:(iqdata_one_file.shape[0]//block_length)*block_length], (iqdata_one_file.shape[0]//block_length,block_length)) # discard the extra elements for uneven array
# print("After: ", iqdata_one_file.shape)
iqdata_one_file = np.expand_dims(iqdata_one_file,axis=2)
iqdata_one_file = np.concatenate([iqdata_one_file['raw-iq'].real,
iqdata_one_file['raw-iq'].imag], axis=2)
if count == 0:
inputs = iqdata_one_file
# print("coming here..", inputs.shape)
else:
inputs = np.concatenate((inputs, iqdata_one_file), axis=0)
# Creating the labels: 2 Label version
if num_classes == 2:
one_file_label = np.zeros((iqdata_one_file.shape[0], num_classes))
if 'LTE_DSSS' in filename:
one_file_label[:, 1] = 1
lte_dsss_count = lte_dsss_count + 1
else:
one_file_label[:, 0] = 1
only_lte_count = only_lte_count + 1
if count == 0:
labels = one_file_label
# print("coming here too..", labels.shape)
else:
labels = np.concatenate((labels, one_file_label), axis=0)
# 7-label version
else:
one_file_label = read_and_extract_metadata_NEU_dataset(metadata_path, meta_filename, iqdata_one_file.shape[0], num_classes)
if count == 0:
labels = one_file_label
# print("coming here too..", labels.shape)
else:
labels = np.concatenate((labels, one_file_label), axis=0)
# Counting the files:
if 'LTE_DSSS' in filename:
lte_dsss_count = lte_dsss_count + 1
else:
only_lte_count = only_lte_count + 1
# total files
count = count + 1
# display DataFrame
print("Input shape: ", inputs.shape)
print("Label shape: ", labels.shape)
print("TOTAL READ FILES: ", count)
print("TOTAL LTE FILES: ", only_lte_count)
print("TOTAL LTE and DSSS FILES: ", lte_dsss_count)
return inputs, labels
def generate_inputs_labels_IQ_for_NWRA_dataset(feature_path, blocks = [131072], block_length = 256, num_classes=2, snr_list = [0, 5, 10], sir_list=[0, 5, 10], dsss_type ='all', percentage_to_read = 0.1, slicing = True, slice_length = 2048):
inputs = np.zeros(1)
labels = np.zeros(1)
# metadata_path = feature_path + 'NWRA_data/Metadata/'
iq_path = feature_path + 'NWRA_data/IQ/'
# iq_path = feature_path + 'test/IQ/'
count = 0
only_lte_count = 0
lte_dsss_count = 0
dtype_all = scipy.dtype([('raw-iq', scipy.complex64)]) # gr_complex is '32fc' --> make any sense?
# random_number = random.random()
for filename in os.listdir(iq_path):
# print("THe length of file", filename, " is : ", len(filename.split('_')))
if len(filename.split('_')) <= 6:
continue
if random.random() > percentage_to_read:
# print("Continuing, ", filename)
continue
# print("reading file: ", filename)
with open(iq_path+filename, mode='rb') as file: # b is important -> binary
iqdata_one_file = scipy.fromfile(file, dtype=dtype_all)
iqdata_one_file = iqdata_one_file[2:] # skipping first two elements (as per Chad's binary.m file)
if slicing == True:
# print("Shapes: ", iqdata_one_file.shape)
slice_index = random.randint(0, iqdata_one_file.shape[0] - slice_length - 1)
iqdata_one_file = iqdata_one_file[slice_index:slice_index + slice_length]
iqdata_one_file = np.expand_dims(iqdata_one_file, axis=1)
iqdata_one_file = np.transpose(iqdata_one_file)
else: # reading the whole i/q file stacked by different block lengths (block_len)
iqdata_one_file = np.reshape(iqdata_one_file[:(iqdata_one_file.shape[0]//block_length)*block_length], (iqdata_one_file.shape[0]//block_length,block_length)) # discard the extra elements for uneven array
# print("After: ", iqdata_one_file.shape)
iqdata_one_file = np.expand_dims(iqdata_one_file,axis=2)
iqdata_one_file = np.concatenate([iqdata_one_file['raw-iq'].real,
iqdata_one_file['raw-iq'].imag], axis=2)
one_file_label = np.zeros((iqdata_one_file.shape[0], num_classes))
# Creating the labels: 2 Label version
if num_classes == 2:
flag = extract_metadata_from_NWRA_dataset(filename, dsss_type)
if 'SKIP' in flag:
continue
elif 'LTE_DSSS' in flag:
one_file_label[:, 1] = 1
lte_dsss_count = lte_dsss_count + 1
else:
one_file_label[:, 0] = 1
only_lte_count = only_lte_count + 1
if count == 0:
inputs = iqdata_one_file
labels = one_file_label
# print("coming here too..", labels.shape)
else:
inputs = np.concatenate((inputs, iqdata_one_file), axis=0)
labels = np.concatenate((labels, one_file_label), axis=0)
print("printing the shapes for ", filename, " is: ", iqdata_one_file.shape, one_file_label.shape)
# if iqdata_one_file.shape[0]!= one_file_label.shape[0]:
# print("!!!!MISMATCH!!!!")
count = count + 1
# display DataFrame
print("Input shape: ", inputs.shape)
print("Label shape: ", labels.shape)
print("TOTAL READ FILES: ", count)
print("TOTAL LTE FILES: ", only_lte_count)
print("TOTAL LTE and DSSS FILES: ", lte_dsss_count)
return inputs, labels
def generate_inputs_labels_NC_C_for_NEU_dataset(feature_path, feature_type ='nc', feature_options=[0], blocks = [131072], num_classes=2, snr_list = [0, 5, 10], sir_list=[0, 5, 10], strategy = 0):
"""
This function is to load non-conjugate features in a specific session at a particular time.
:param feature_path: str, the file path for the feature
:param feature_type: str, the feature type: 'nc' or 'c'
"""
inputs = np.zeros(1)
labels =np.zeros(1)
input_label_dic = defaultdict(list)
metadata_path = feature_path + 'IQDataSet_LTE_DSSS_v2/Metadata/'
feature_path = feature_path + 'NEU_LTE_DSSS_Dataset_2_CSP/'
# NWRA_data\131072 'test/CSP/'
# feature_path = feature_path + 'NWRA_data/131072/'
count = 0
only_lte_count = 0
lte_dsss_count = 0
for filename in os.listdir(feature_path):
sir_for_this_file = sir_list[0] # intitlizing the SIR values for only LTE signals
# print("**********At Starting: ", filename, sir_for_this_file, sir_list)
file_extension = '.NC'
if feature_type == 'c':
file_extension = '.C'
if file_extension in filename and any(str(ext) in filename for ext in blocks):
# creating the features
# print('File with the features:', filename)
meta_filename = filename[::-1].split('_',2)[-1][::-1]
# print("meta file name: ", meta_filename)
snr_for_this_file = extract_snr_from_metadata_NEU_dataset(metadata_path, meta_filename)
if 'LTE_DSSS' in filename:
sir_for_this_file = extract_sir_from_metadata_NEU_dataset(metadata_path, meta_filename)
# print("THE SIR VALUES: ", snr_for_this_file.item(), sir_list)
# print("**********Before Entering: ", filename, sir_for_this_file, sir_list, snr_for_this_file, snr_list)
if ((snr_for_this_file in snr_list) and (sir_for_this_file in sir_list)): # the second part of the condition is not being used for 'only lte' signals
# print("**********Entering: ", filename, sir_for_this_file, sir_list, snr_for_this_file, snr_list)
try:
one_file_df = pd.read_csv(feature_path+filename, sep=" ", header=None).dropna(axis=1, how='all')
column_index = list(one_file_df.columns)
selected_columns = [column_index[i] for i in feature_options]
# print("one_df and input shape:", one_file_df[selected_columns].to_numpy().shape, inputs.shape)
one_file_np = one_file_df[selected_columns].to_numpy()
except pd.errors.EmptyDataError:
# print("Note:", filename, "was empty. Skipping.")
# count = count - 1 # decreasing the total file count
# one_file_df =
one_file_np = np.zeros((1, len(feature_options)))
# continue
if strategy == 1:
one_file_np = organize_CSP_features_2D(one_file_np)
if strategy == 2:
one_file_np = extract_statistic_from_CSP_features(one_file_np)
one_file_np = np.expand_dims(one_file_np, axis=0)
if strategy == 3:
one_file_np = organize_CSP_features_3D(one_file_np)
if strategy == 4:
one_file_np = extract_statistic_from_CSP_nc_features_one_col(one_file_np)
one_file_np = np.expand_dims(one_file_np, axis=0)
if count == 0:
inputs = one_file_np
# print("coming here..", inputs.shape)
else:
inputs = np.concatenate((inputs, one_file_np), axis=0)
one_file_label = np.zeros((one_file_np.shape[0], num_classes))
# if strategy == 1: one_file_label = np.zeros((1, num_classes))
# Creating the labels: 2 Label version
if num_classes == 2:
if 'LTE_DSSS' in filename:
# if strategy == 1: one_file_label[1] = 1
one_file_label[:, 1] = 1
lte_dsss_count = lte_dsss_count +1
else:
one_file_label[:, 0] = 1
only_lte_count = only_lte_count + 1
if count == 0:
labels = one_file_label
# print("coming here too..", labels.shape)
else:
labels = np.concatenate((labels, one_file_label), axis=0)
# 7-label version
else:
one_file_label = read_and_extract_metadata_NEU_dataset(metadata_path, meta_filename, one_file_np.shape[0], num_classes)
if count == 0:
labels = one_file_label
# print("coming here too..", labels.shape)
else:
labels = np.concatenate((labels, one_file_label), axis=0)
#Counting the files:
if 'LTE_DSSS' in filename:
lte_dsss_count = lte_dsss_count +1
else:
only_lte_count = only_lte_count + 1
#total files
count = count + 1
# adding the input and labels to the dictionary
input_label_dic['input'].append(one_file_np)
input_label_dic['label'].append(one_file_label[0])
# display DataFrame
print("Input shape: ", inputs.shape)
print("Label shape: ", labels.shape)
print("TOTAL READ FILES: ", count)
print("TOTAL LTE FILES: ", only_lte_count)
print("TOTAL LTE and DSSS FILES: ", lte_dsss_count)
print("***** THE SIZE OF THE DICTIONARY INPUT AND LABELS: ", len(input_label_dic['input']), len(input_label_dic['label']))
return inputs, labels, input_label_dic
def generate_inputs_labels_NC_C_for_NWRA_dataset(feature_path, feature_type ='nc', feature_options=[0], blocks = [131072], num_classes = 2, strategy = 0, dsss_type ='all'):
"""
This function is to load non-conjugate features in a specific session at a particular time.
:param feature_path: str, the file path for the feature
:param feature_type: str, the feature type: 'nc' or 'c'
"""
inputs = np.zeros(1)
labels =np.zeros(1)
input_label_dic = defaultdict(list)
feature_path = feature_path + 'NWRA_data/'+str(blocks[0])+'/'
count = 0
only_lte_count = 0
lte_dsss_count = 0
for filename in os.listdir(feature_path):
file_extension = '.NC'
if feature_type == 'c':
file_extension = '.C'
if file_extension in filename and any(str(ext) in filename for ext in blocks):
# block_number =
# print("Reading...", filename)
try:
one_file_df = pd.read_csv(feature_path+filename, sep=" ", header=None).dropna(axis=1, how='all')
column_index = list(one_file_df.columns)
selected_columns = [column_index[i] for i in feature_options]
# print("one_df and input shape:", one_file_df[selected_columns].to_numpy().shape, inputs.shape)
one_file_np = one_file_df[selected_columns].to_numpy()
except pd.errors.EmptyDataError:
# print("Note:", filename, "was empty. Skipping.")
# count = count - 1 # decreasing the total file count
# one_file_df =
one_file_np = np.zeros((1, len(feature_options)))
# continue
if strategy == 1:
one_file_np = organize_CSP_features_2D(one_file_np)
if strategy == 2:
one_file_np = extract_statistic_from_CSP_features(one_file_np)
one_file_np = np.expand_dims(one_file_np, axis=0)
if strategy == 3:
one_file_np = organize_CSP_features_3D(one_file_np)
if strategy == 4:
one_file_np = extract_statistic_from_CSP_nc_features_one_col(one_file_np) # changed for comjugate feautures
one_file_np = np.expand_dims(one_file_np, axis=0)
one_file_label = np.zeros((one_file_np.shape[0], num_classes))
# Creating the labels: 2 Label version
if num_classes == 2:
flag = extract_metadata_from_NWRA_dataset(filename, dsss_type)
if 'SKIP' in flag:
continue