-
Notifications
You must be signed in to change notification settings - Fork 1
/
signal_processing.py
304 lines (252 loc) · 11 KB
/
signal_processing.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
# creates a dataset of the signals to analyze with different transformations
# work in progress
import math
from os.path import exists
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import scipy.stats, scipy.signal
import scipy.stats, scipy.signal
# TODO: not seeing the functions giving good results, -> should we plot more...
# read a bunch in
from sklearn.discriminant_analysis import LinearDiscriminantAnalysis
from sklearn.ensemble import RandomForestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split, StratifiedKFold, cross_val_score
from sklearn.naive_bayes import GaussianNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.neural_network import MLPClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier
names2 = ['chris', 'daniel','eva', 'judah', 'leo', 'quinn']
actions2 = ['blink', 'bite']
time = ['5seconds']
filenames, blinkfiles, bitefiles = [], [], []
for i in names2:
for j in actions2:
if j=='blink':
for k in time:
blinkfiles.append(i + "_" + j + '_'+ k + ".csv")
else:
for k in time:
bitefiles.append(i + "_" + j + "_"+ k + ".csv")
dates = ['02-24-22', '03-24-22', '03-24']
names3 = ['eva', 'chris', 'leo', 'judah']
times = ['6seconds', '5seconds', '10seconds', '8seconds']
actions3 = ['blink', 'bite']
for i in names3:
for j in actions3:
if j=='blink':
for k in times:
for d in dates:
blinkfiles.append(d + "_" +i + "_" + j + '_'+ k + ".csv")
else:
for k in times:
for d in dates:
bitefiles.append(d + "_" + i + "_" + j + "_"+ k + ".csv")
newfiles = [
'03-16-22-2_mark_blink_6seconds.csv',
'03-16-22_dan_blink_6seconds.csv',
'03-16-22_eva_blink_6seconds.csv',
'03-16-22_judah_blink_6seconds.csv',
'03-16-22_leo_blink_6seconds.csv',
'03-16-22_mark_blink_6seconds.csv',
'03-16-22_mark_blink_6seconds.csv'
]
for i in newfiles:
blinkfiles.append(i)
bites = ['03-16-22_dan_bite_6seconds.csv',
'03-16-22_eva_bite_6seconds.csv',
'03-16-22_eva_bite_10seconds.csv',
'03-16-22_judah_bite_6seconds.csv',
'03-16-22_leo_bite_6seconds.csv',
'03-16-22_mark_bite_6seconds.csv',"03-16-22-2_mark_bite_6seconds.csv"]
for i in bites:
bitefiles.append(i)
eyes = ["03-16-22-2_mark_eyebrow_6seconds.csv",
'03-16-22_mark_eyebrow_6seconds.csv'
'03-16-22_leo_eyebrow_6seconds.csv',
'03-16-22_judah_eyebrow_6seconds.csv',
]
df = pd.DataFrame(columns = ['signal', 'channel', 'action', 'accumulated energy', 'initial steepness', 'ascent steppness', 'decsent steepness', 'meanvariance', 'snr'])
prefix = 'Brainwaves-to-Letters/Bite and Blink Data Analysis v1/'
for blink in blinkfiles:
if(exists(prefix+blink)):
d = pd.read_csv(prefix + blink)
darray = np.array(d.c3)
df = df.append({'signal' : darray, 'channel' : 3, 'action' : 0}, ignore_index = True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 0}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 0}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 0}, ignore_index = True)
elif(exists(blink)):
d = pd.read_csv(blink)
darray = np.array(d.c3)
df = df.append({'signal' : darray, 'channel' : 3,'action' : 0}, ignore_index = True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 0}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 0}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 0}, ignore_index = True)
elif(exists('Brainwaves-to-Letters/' + blink)):
d = pd.read_csv('Brainwaves-to-Letters/' + blink)
darray = np.array(d.c3)
df = df.append({'signal': darray, 'channel' : 3,'action': 0}, ignore_index=True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 0}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 0}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 0}, ignore_index = True)
for bite in bitefiles:
if(exists(prefix+bite)):
d = pd.read_csv(prefix + bite)
darray = np.array(d.c3)
df = df.append({'signal' : darray, 'channel' : 3,'action' : 1}, ignore_index = True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 1}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 1}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 1}, ignore_index = True)
elif(exists(bite)):
d = pd.read_csv(bite)
darray = np.array(d.c3)
df = df.append({'signal' : darray, 'channel' : 3,'action' : 1}, ignore_index = True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 1}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 1}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 1}, ignore_index = True)
elif(exists('Brainwaves-to-Letters/' + bite)):
d = pd.read_csv('Brainwaves-to-Letters/' + bite)
darray = np.array(d.c3)
df = df.append({'signal': darray, 'channel' : 3,'action': 1}, ignore_index=True)
darray = np.array(d.c1)
df = df.append({'signal' : darray, 'channel' : 1, 'action' : 1}, ignore_index = True)
darray = np.array(d.c2)
df = df.append({'signal' : darray, 'channel' : 2, 'action' : 1}, ignore_index = True)
darray = np.array(d.c4)
df = df.append({'signal' : darray, 'channel' : 4, 'action' : 1}, ignore_index = True)
# for signal, transform, append under transform coloumn
def AccumulatedEnergy(event):
sq = [x ** 2 for x in event]
energy = sum(sq)
return energy
def InitialSteepness(event, samplingInterval): # changed x to be event
# Not sure how many samples are in a typical event
start = event[: math.floor(len(event) / 2)]
slope = []
for i in range(len(event) - 1):
slope.append((event[i] - event[i + 1]) / samplingInterval)
avgSlope = sum(slope) / len(slope)
return avgSlope
def AscentSteepness(event, samplingInterval):
# Not sure how many samples are in a typical event
cutoff = event.index(max(event))
start = event[: cutoff]
slope = []
for i in range(len(x) - 1):
slope.append(abs((x[i] - x[i + 1]) / samplingInterval))
avgSlope = sum(slope) / len(slope)
return avgSlope
def DescentSteepness(event, samplingInterval):
# Not sure how many samples are in a typical event
cutoff = event.index(max(event))
start = event[cutoff:]
slope = []
for i in range(len(x) - 1):
slope.append(abs((x[i] - x[i + 1]) / samplingInterval))
avgSlope = sum(slope) / len(slope)
return avgSlope
def MeanVarRatio(event):
ratio = []
lenSS = 20
numSubsections = math.floor(len(event) / lenSS)
for i in range(numSubsections):
ratio.append((sum(event[(lenSS * i):(lenSS * (i + 1))]) / lenSS) / np.var(event[(lenSS * i):(lenSS * (i + 1))]))
m = max(ratio)
return m
def MeanScaledVar(event, lenSS=20):
ratio = []
numSubsections = math.floor(len(event) / lenSS)
for i in range(numSubsections):
ratio.append((sum(event[(lenSS * i):(lenSS * (i + 1))]) / lenSS) * np.var(event[(lenSS * i):(lenSS * (i + 1))]))
m = max(ratio)
return m
def signaltonoise(a, axis, ddof):
a = np.asanyarray(a)
m = a.mean(axis)
sd = a.std(axis=axis, ddof=ddof)
return np.where(sd == 0, 0, m / sd)
# this has format issues,
# also not really seeing a trend but should try with other electrodes and a shortened sample 1s ish invervals over the action
for s in df.signal:
df = df.append({'SNR' : signaltonoise(s[3*256:4*256],0,0)}, ignore_index=True)
#fix this
df.SNR[0:11] = df.SNR[12:23]
df.SNR[11] = df.SNR[23]
df = df.drop([12,13,14,15,16,17,18,19,20,21,22,23])
snrs, m1, m2, m3, m4, m5, m6 = [], [], [], [], [], [], []
for s in df.signal:
snrs.append(signaltonoise(s,0,0))
m1.append(AccumulatedEnergy(s))
m2.append(MeanVarRatio(s))
m3.append(MeanScaledVar(s, lenSS=20))
#'accumulated energy', 'initial steepness', 'ascent steppness', 'decsent steepness', 'meanvariance', 'snr'
df['snr'] = snrs
df['accumulated energy'] = m1
df['meanvariance'] = m2
df['mean scaled var'] = m3
for i in range(len(df.signal)):
if (df.action[i] == 0):
plt.plot(df.signal[i], color = 'blue', label = 'blink')
else:
plt.plot(df.signal[i], color = 'green', label = 'bite')
#TODO: make a time labeling application
#TODO: just see if the NN can handle any of these, some stats maybe first
# we only want one signal in here ^ so some step to choose the best electrode, or we use a few of them
# analytics
x, y = [],[]
for i in range(len(df.signal)):
x.append(df.signal[i])
y.append(df.action[i])
X_train, X_validation, Y_train, Y_validation = train_test_split(x, y, test_size=0.2, random_state=42)
models = []
models.append(('LR', LogisticRegression(solver='liblinear', multi_class='ovr')))
models.append(('LDA', LinearDiscriminantAnalysis()))
models.append(('KNN',KNeighborsClassifier()))
models.append(('CART', DecisionTreeClassifier()))
models.append(('NB', GaussianNB()))
models.append(('SVM', SVC(gamma='auto')))
models.append(('RF', RandomForestClassifier()))
models.append(('MLP', MLPClassifier(hidden_layer_sizes=(64,15))))
# evaluate each model
results = []
names = []
for name, model in models:
kfold = StratifiedKFold(n_splits=9, random_state=1, shuffle=True)
cv_results = cross_val_score(model, X_train, Y_train, cv=kfold, scoring='accuracy')
results.append(cv_results)
names.append(name)
print('%s: %f (%f)' % (name, cv_results.mean(), cv_results.std()))
# Visualizations
f1 = 'Brainwaves-to-Letters/02-24-22_eva_bite_8seconds'
f2 = 'Brainwaves-to-Letters/02-24-22_eva_blink_5seconds.csv'
f3 = 'Brainwaves-to-Letters/02-24-22_eva_blink_6seconds.csv'
f4 = 'Brainwaves-to-Letters/eva_bite_10seconds'
data = pd.read_csv(f1)
data = data.drop(columns='Unnamed: 0')
data.time = data.time - data.time[0]
data.time = data.time/1000000 # into seconds
#cut the data actiontime - 1: actiontime + 2, sfreq = 256 so 256*starts:256*ends
# 5or6s 2:6, 10s 4:8, 8s 3:7
# static time series
fig = plt.plot(data.time[256*3:256*7], data.c3[256*3:256*7], linewidth=0.1)
# massive spike in c3 only
#static square
fig = plt.plot(data.time[256*3:256*7], (data.c3[256*3:256*7])**2, linewidth=0.1)