-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdriving_transit_acc_magnet.py
226 lines (184 loc) · 7.22 KB
/
driving_transit_acc_magnet.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
import pandas as pd
import math as math
import glob
from h2o import h2o
# Loading Accelerometer Data
drivingCarPath = r'/Users/yalcin.yenigun/dev/workspaces/gsu/Sample Dataset/Driving-Transit/Driving Car/csv'
drivingCarFiles = glob.glob(drivingCarPath + "/0_Accelerometer*.csv")
accDfCar = pd.DataFrame()
list_ = []
for file_ in drivingCarFiles:
df = pd.read_csv(file_, index_col=None, header=0)
list_.append(df)
accDfCar = pd.concat(list_)
accDfCar['label'] = 'driving car'
transitPath = r'/Users/yalcin.yenigun/dev/workspaces/gsu/Sample Dataset/Driving-Transit/Transit/csv'
transitFiles = glob.glob(transitPath + "/0_Accelerometer*.csv")
accDfTransit = pd.DataFrame()
list_ = []
for file_ in transitFiles:
df = pd.read_csv(file_, index_col=None, header=0)
list_.append(df)
accDfTransit = pd.concat(list_)
accDfTransit['label'] = 'transit'
accDfFrames = [accDfCar, accDfTransit]
accDf = pd.concat(accDfFrames)
xV = (accDf['x'] * accDf['x']) + (accDf['y'] * accDf['y']) + (accDf['z'] * accDf['z'])
accDf['acc_mag'] = xV
accDf['acc_mag'] = accDf['acc_mag'].apply(math.sqrt)
# Convert timestamp to date time
accDf['start'] = pd.to_datetime(accDf['start'])
accDf['end'] = pd.to_datetime(accDf['end'])
print(len(accDf.index))
# Windowing Accelerometer Data
accMagFeatures = pd.DataFrame()
accDf['timestamps'] = pd.to_datetime(accDf['timestamps'])
accDf.set_index(['timestamps'])
accDf = accDf.sort_values(by='timestamps')
accMagDf = accDf[['timestamps', 'start', 'acc_mag', 'x', 'y', 'z', 'label']]
accMagDf = accMagDf.set_index(['timestamps'])
accMagFeatures['acc_mag_mean'] = accMagDf['acc_mag'].rolling('1s').mean()
accMagFeatures['acc_mag_std'] = accMagDf['acc_mag'].rolling('1s').std()
accMagFeatures['acc_mag_var'] = accMagDf['acc_mag'].rolling('1s').var()
accMagFeatures['acc_mag_min'] = accMagDf['acc_mag'].rolling('1s').min()
accMagFeatures['acc_mag_max'] = accMagDf['acc_mag'].rolling('1s').max()
accMagFeatures['acc_x_mean'] = accMagDf['x'].rolling('1s').mean()
accMagFeatures['acc_x_std'] = accMagDf['x'].rolling('1s').std()
accMagFeatures['acc_x_var'] = accMagDf['x'].rolling('1s').var()
accMagFeatures['acc_x_min'] = accMagDf['x'].rolling('1s').min()
accMagFeatures['acc_x_max'] = accMagDf['x'].rolling('1s').max()
accMagFeatures['acc_y_mean'] = accMagDf['y'].rolling('1s').mean()
accMagFeatures['acc_y_std'] = accMagDf['y'].rolling('1s').std()
accMagFeatures['acc_y_var'] = accMagDf['y'].rolling('1s').var()
accMagFeatures['acc_y_min'] = accMagDf['y'].rolling('1s').min()
accMagFeatures['acc_y_max'] = accMagDf['y'].rolling('1s').max()
accMagFeatures['acc_z_mean'] = accMagDf['z'].rolling('1s').mean()
accMagFeatures['acc_z_std'] = accMagDf['z'].rolling('1s').std()
accMagFeatures['acc_z_var'] = accMagDf['z'].rolling('1s').var()
accMagFeatures['acc_z_min'] = accMagDf['z'].rolling('1s').min()
accMagFeatures['acc_z_max'] = accMagDf['z'].rolling('1s').max()
accMagFeatures['label'] = accMagDf['label']
# Loading Magnetometer Data
drivingCarMagFiles = glob.glob(drivingCarPath + "/0_Magnetometer*.csv")
drivingCarMagDf = pd.DataFrame()
list_ = []
for file_ in drivingCarMagFiles:
df = pd.read_csv(file_, index_col=None, header=0)
list_.append(df)
drivingCarMagDf = pd.concat(list_)
drivingCarMagDf['label'] = 'driving car'
transitMagFiles = glob.glob(transitPath + "/0_Magnetometer*.csv")
transitMagDf = pd.DataFrame()
list_ = []
for file_ in transitMagFiles:
df = pd.read_csv(file_, index_col=None, header=0)
list_.append(df)
transitMagDf = pd.concat(list_)
transitMagDf['label'] = 'transit'
magDfFrames = [drivingCarMagDf, transitMagDf]
magDf = pd.concat(magDfFrames)
# Windowing Magnetometer Data
magV = (magDf['x'] * magDf['x']) + (magDf['y'] * magDf['y']) + (magDf['z'] * magDf['z'])
magDf['mag_magnitude'] = magV
magDf['mag_magnitude'] = magDf['mag_magnitude'].apply(math.sqrt)
# Convert timestamp to date time
magDf['start'] = pd.to_datetime(magDf['start'])
magDf['end'] = pd.to_datetime(magDf['end'])
print(len(magDf.index))
magFeatures = pd.DataFrame()
magDf['timestamps'] = pd.to_datetime(magDf['timestamps'])
magDf.set_index(['timestamps'])
magDf = magDf.sort_values(by='timestamps')
magDf = magDf[['timestamps', 'start', 'mag_magnitude', 'x', 'y', 'z', 'label']]
magDf = magDf.set_index(['timestamps'])
magFeatures['mag_mag_mean'] = magDf['mag_magnitude'].rolling('1s').mean()
magFeatures['mag_mag_std'] = magDf['mag_magnitude'].rolling('1s').std()
magFeatures['mag_mag_var'] = magDf['mag_magnitude'].rolling('1s').var()
magFeatures['mag_mag_min'] = magDf['mag_magnitude'].rolling('1s').min()
magFeatures['mag_mag_max'] = magDf['mag_magnitude'].rolling('1s').max()
magFeatures['mag_x_mean'] = magDf['x'].rolling('1s').mean()
magFeatures['mag_x_std'] = magDf['x'].rolling('1s').std()
magFeatures['mag_x_var'] = magDf['x'].rolling('1s').var()
magFeatures['mag_x_min'] = magDf['x'].rolling('1s').min()
magFeatures['mag_x_max'] = magDf['x'].rolling('1s').max()
magFeatures['mag_y_mean'] = magDf['y'].rolling('1s').mean()
magFeatures['mag_y_std'] = magDf['y'].rolling('1s').std()
magFeatures['mag_y_var'] = magDf['y'].rolling('1s').var()
magFeatures['mag_y_min'] = magDf['y'].rolling('1s').min()
magFeatures['mag_y_max'] = magDf['y'].rolling('1s').max()
magFeatures['mag_z_mean'] = magDf['z'].rolling('1s').mean()
magFeatures['mag_z_std'] = magDf['z'].rolling('1s').std()
magFeatures['mag_z_var'] = magDf['z'].rolling('1s').var()
magFeatures['mag_z_min'] = magDf['z'].rolling('1s').min()
magFeatures['mag_z_max'] = magDf['z'].rolling('1s').max()
magFeatures['label'] = magDf['label']
# Merge Features
accMagFeatures['timestamps'] = accMagFeatures.index
magFeatures['timestamps'] = magFeatures.index
accMagFeatures = accMagFeatures.sort_values(by='timestamps')
magFeatures = magFeatures.sort_values(by='timestamps')
allFeatures = pd.merge_asof(accMagFeatures, magFeatures, on='timestamps', by='label',
tolerance=pd.Timedelta('100ms'))
# Remove Null Features
print(len(accMagFeatures.index))
print(len(magFeatures.index))
print(len(allFeatures.index))
allFeatures = allFeatures[~(allFeatures.mag_mag_mean.isnull()) & ~(allFeatures.acc_mag_mean.isnull())]
len(allFeatures.index)
# Run Classification For Magnetometer & Accelerometer
h2o.init()
h2o.remove_all()
allFeatures = h2o.H2OFrame(allFeatures)
continuous_feature_columns = [
'mag_mag_mean',
'mag_mag_std',
'mag_mag_var',
'mag_mag_min',
'mag_mag_max',
'mag_x_mean',
'mag_x_std',
'mag_x_var',
'mag_x_min',
'mag_x_max',
'mag_y_mean',
'mag_y_std',
'mag_y_var',
'mag_y_min',
'mag_y_max',
'mag_z_mean',
'mag_z_std',
'mag_z_var',
'mag_z_min',
'mag_z_max',
'acc_mag_mean',
'acc_mag_std',
'acc_mag_var',
'acc_mag_min',
'acc_mag_max',
'acc_x_mean',
'acc_x_std',
'acc_x_var',
'acc_x_min',
'acc_x_max',
'acc_y_mean',
'acc_y_std',
'acc_y_var',
'acc_y_min',
'acc_y_max',
'acc_z_mean',
'acc_z_std',
'acc_z_var',
'acc_z_min',
'acc_z_max'
]
random_forest_model = h2o.H2ORandomForestEstimator(
model_id="DrivingTransitAccelerometerMagnetometer",
ntrees=20,
max_depth=10,
min_rows=4,
nfolds=10,
seed=12345
)
random_forest_model.train(x=continuous_feature_columns,
y='label',
training_frame=allFeatures)