RS2002 commited on
Commit
a122232
·
verified ·
1 Parent(s): 43ae8c5

Upload 6 files

Browse files
WiCount.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d578ba4ec4d64ef7089e0fa5c8b47498a76fce5f7d32fb7c7f7f994235915ae1
3
+ size 9389189
data_process_example/README.md ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # How to Run
2
+
3
+ 1. Execute `process1.py` to convert the `csv` file into a `pkl` file.
4
+
5
+ 2. Run one of the `process2` scripts based on your requirements:
6
+
7
+
8
+
9
+ **Note:** We use `-1000` to represent the position of package loss. You can apply various interpolation methods to fill these gaps. We highly encourage you to try our CSI-BERT model to recover the lost packages. ([CSI-BERT](https://github.com/RS2002/CSI-BERT), [CSI-BERT2](https://github.com/RS2002/CSI-BERT2))
10
+
11
+
12
+
13
+ (1) If you want to process each record into a long sequence, run `process2.py`. You can refer to `dataset.py` in [CSI-BERT2](https://github.com/RS2002/CSI-BERT2) for guidance.
14
+
15
+
16
+
17
+ (2) If you prefer to split each record into multiple fixed-length samples, run `process2-split.py` and modify the `length` parameter in the code to your desired length. You can refer to `dataset.py` in [CSI-BERT](https://github.com/RS2002/CSI-BERT), [CrossFi](https://github.com/RS2002/CrossFi), [KNN-MMD](https://github.com/RS2002/CrossFi), and [LoFi](https://github.com/RS2002/LoFi/tree/main/network_examples) for usage instructions.
18
+
19
+
20
+
21
+ (3) `process2-squeeze-split.py` functions similarly to `process2-split.py`, but it excludes all lost packages.
data_process_example/process1.py ADDED
@@ -0,0 +1,50 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ import pickle
3
+ import os
4
+ import pandas as pd
5
+
6
+ root="./data/"
7
+ data=[]
8
+ csi_vaid_subcarrier_index = range(0, 52)
9
+
10
+ def handle_complex_data(x, valid_indices):
11
+ real_parts = []
12
+ imag_parts = []
13
+ for i in valid_indices:
14
+ real_parts.append(x[i * 2])
15
+ imag_parts.append(x[i * 2 - 1])
16
+ return np.array(real_parts) + 1j * np.array(imag_parts)
17
+
18
+
19
+ for people_num in os.listdir(root):
20
+ if len(people_num)>1:
21
+ continue
22
+ print(people_num)
23
+ path=os.path.join(root,people_num)
24
+
25
+ for file in os.listdir(path):
26
+ if file[-3:] != "csv":
27
+ continue
28
+ print(file)
29
+ df = pd.read_csv(os.path.join(path,file))
30
+ df.dropna(inplace=True)
31
+ df['data'] = df['data'].apply(lambda x: eval(x))
32
+ complex_data = df['data'].apply(lambda x: handle_complex_data(x, csi_vaid_subcarrier_index))
33
+ magnitude = complex_data.apply(lambda x: np.abs(x))
34
+ phase = complex_data.apply(lambda x: np.angle(x, deg=True))
35
+ time = np.array(df['timestamp'])
36
+ local_time = np.array(df['local_timestamp'])
37
+
38
+ data.append({
39
+ 'csi_time':time,
40
+ 'csi_local_time':local_time,
41
+ 'people_num': eval(people_num),
42
+ 'magnitude': np.array([np.array(a) for a in magnitude]),
43
+ 'phase': np.array([np.array(a) for a in phase]),
44
+ 'CSI': np.array([np.array(a) for a in complex_data])
45
+ })
46
+
47
+ # 保存全局字典为一个pickle文件
48
+ output_file = './csi_data.pkl'
49
+ with open(output_file, 'wb') as f:
50
+ pickle.dump(data, f)
data_process_example/process2-split.py ADDED
@@ -0,0 +1,97 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ import pickle
3
+ import copy
4
+
5
+ def get_time(s):
6
+ s=s.split()[-1]
7
+ s=s.split(":")
8
+ h=float(s[0])
9
+ m=float(s[1])
10
+ t=float(s[2])
11
+ total=h*3600+m*60+t
12
+ return h,m,t,total
13
+
14
+ gap=1
15
+ length=gap*100
16
+ pad=[-1000]*52
17
+ action_list=[]
18
+ people_list=[]
19
+ timestamp=[]
20
+ magnitudes=[]
21
+ phases=[]
22
+ loacl_gap=10000
23
+
24
+
25
+ with open("./csi_data.pkl", 'rb') as f:
26
+ csi = pickle.load(f)
27
+
28
+ for data in csi:
29
+ csi_time=data['csi_time']
30
+ local_time=data['csi_local_time']
31
+ magnitude=data['magnitude']
32
+ phase=data['phase']
33
+ people=data['people_num']
34
+ action=people
35
+ start_time=None
36
+ last_local=None
37
+ current_magnitude=[]
38
+ current_phase=[]
39
+ current_timestamp=[]
40
+ for i in range(len(csi_time)):
41
+ _, _, _, current_time = get_time(csi_time[i])
42
+ if start_time is None or current_time-start_time>gap:
43
+ if start_time is not None:
44
+ if len(current_magnitude)>=length:
45
+ current_magnitude=current_magnitude[:length]
46
+ current_phase=current_phase[:length]
47
+ current_timestamp=current_timestamp[:length]
48
+ else:
49
+ add=length-len(current_magnitude)
50
+ delta=(current_timestamp[0]+length*loacl_gap-current_timestamp[-1])/add
51
+ for j in range(add):
52
+ current_magnitude.append(pad)
53
+ current_phase.append(pad)
54
+ current_timestamp.append(current_timestamp[-1]+delta)
55
+ magnitudes.append(copy.deepcopy(current_magnitude))
56
+ phases.append(copy.deepcopy(current_phase))
57
+ timestamp.append(copy.deepcopy(current_timestamp))
58
+ action_list.append(action)
59
+ people_list.append(people)
60
+ current_magnitude = []
61
+ current_phase = []
62
+ current_timestamp = []
63
+ start_time=current_time
64
+ last_local=local_time[i]
65
+ current_magnitude.append(magnitude[i])
66
+ current_phase.append(phase[i])
67
+ current_timestamp.append(local_time[i])
68
+ else:
69
+ local = local_time[i]
70
+ num=round((local-last_local-loacl_gap)/loacl_gap)
71
+ if num>0:
72
+ delta=(local-last_local)/(num+1)
73
+ for j in range(num):
74
+ current_magnitude.append(pad)
75
+ current_phase.append(pad)
76
+ current_timestamp.append(current_timestamp[-1] + delta)
77
+ current_magnitude.append(magnitude[i])
78
+ current_phase.append(phase[i])
79
+ current_timestamp.append(local_time[i])
80
+ last_local=local
81
+
82
+
83
+ action_list=np.array(action_list)
84
+ people_list=np.array(people_list)
85
+ timestamp=np.array(timestamp)
86
+ magnitudes=np.array(magnitudes)
87
+ phases=np.array(phases)
88
+ print(action_list.shape)
89
+ print(people_list.shape)
90
+ print(timestamp.shape)
91
+ print(magnitudes.shape)
92
+ print(phases.shape)
93
+ np.save("./magnitude.npy", np.array(magnitudes))
94
+ np.save("./phase.npy", np.array(phases))
95
+ np.save("./action.npy", np.array(action_list))
96
+ np.save("./people.npy", np.array(people_list))
97
+ np.save("./timestamp.npy", np.array(timestamp))
data_process_example/process2-squeeze-split.py ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ import pickle
3
+ import copy
4
+
5
+
6
+ gap=1
7
+ length=gap*100
8
+ pad=[-1000]*52
9
+ action_list=[]
10
+ people_list=[]
11
+ timestamp=[]
12
+ magnitudes=[]
13
+ phases=[]
14
+ loacl_gap=10000
15
+
16
+
17
+ with open("./csi_data.pkl", 'rb') as f:
18
+ csi = pickle.load(f)
19
+
20
+ for data in csi:
21
+ csi_time=data['csi_time']
22
+ local_time=data['csi_local_time']
23
+ magnitude=data['magnitude']
24
+ phase=data['phase']
25
+ people=data['people_num']
26
+ action=people
27
+
28
+ index=0
29
+ while index<len(magnitude)-length:
30
+ current_magnitude=magnitude[index:index+length]
31
+ current_phase=phase[index:index+length]
32
+ current_timestamp=local_time[index:index+length]
33
+ index+=(length+gap-1)
34
+ magnitudes.append(copy.deepcopy(current_magnitude))
35
+ phases.append(copy.deepcopy(current_phase))
36
+ timestamp.append(copy.deepcopy(current_timestamp))
37
+ action_list.append(action)
38
+ people_list.append(people)
39
+
40
+ action_list=np.array(action_list)
41
+ people_list=np.array(people_list)
42
+ timestamp=np.array(timestamp)
43
+ magnitudes=np.array(magnitudes)
44
+ phases=np.array(phases)
45
+ print(action_list.shape)
46
+ print(people_list.shape)
47
+ print(timestamp.shape)
48
+ print(magnitudes.shape)
49
+ print(phases.shape)
50
+ np.save("./squeeze_data/magnitude.npy", np.array(magnitudes))
51
+ np.save("./squeeze_data/phase.npy", np.array(phases))
52
+ np.save("./squeeze_data/action.npy", np.array(action_list))
53
+ np.save("./squeeze_data/people.npy", np.array(people_list))
54
+ np.save("./squeeze_data/timestamp.npy", np.array(timestamp))
data_process_example/process2.py ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import numpy as np
2
+ import pickle
3
+
4
+
5
+ result=[]
6
+
7
+ pad=[-1000]*52
8
+ loacl_gap=10000
9
+
10
+
11
+ with open("./csi_data.pkl", 'rb') as f:
12
+ csi = pickle.load(f)
13
+
14
+ for data in csi:
15
+ csi_time=data['csi_time']
16
+ local_time=data['csi_local_time']
17
+ magnitude=data['magnitude']
18
+ phase=data['phase']
19
+ people_num=data['people_num']
20
+
21
+ last_local=None
22
+ current_magnitude=[]
23
+ current_phase=[]
24
+ current_timestamp=[]
25
+ for i in range(len(csi_time)):
26
+ if last_local is None:
27
+ last_local=local_time[i]
28
+ current_magnitude.append(magnitude[i])
29
+ current_phase.append(phase[i])
30
+ current_timestamp.append(local_time[i])
31
+ else:
32
+ local = local_time[i]
33
+ num=round((local-last_local-loacl_gap)/loacl_gap)
34
+ if num>0:
35
+ delta=(local-last_local)/(num+1)
36
+ for j in range(num):
37
+ current_magnitude.append(pad)
38
+ current_phase.append(pad)
39
+ current_timestamp.append(current_timestamp[-1] + delta)
40
+ current_magnitude.append(magnitude[i])
41
+ current_phase.append(phase[i])
42
+ current_timestamp.append(local_time[i])
43
+ last_local=local
44
+
45
+ print(len(current_magnitude))
46
+ result.append({
47
+ 'time': np.array(current_timestamp),
48
+ 'action': people_num,
49
+ 'people': people_num,
50
+ 'magnitude': np.array(current_magnitude),
51
+ 'phase': np.array(current_phase)
52
+ })
53
+
54
+ output_file = './data_sequence.pkl'
55
+ with open(output_file, 'wb') as f:
56
+ pickle.dump(result, f)
57
+