-
Notifications
You must be signed in to change notification settings - Fork 1
/
result_visualize.py
206 lines (183 loc) · 7.9 KB
/
result_visualize.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
import numpy as np
import pandas as pd
import utils
from model_config import build_args
import os
import json
import matplotlib.pyplot as plt
plt.rc('font',family='Times New Roman')
plt.rcParams['axes.unicode_minus'] = False
plt.rcParams['pdf.fonttype'] = 42
plt.rcParams['ps.fonttype'] = 42
from pylab import mpl
mpl.rcParams['font.size'] = 18
def temporal_curve():
data_root="./original_data_new"
subjects=["dzh","hzh","xxh","zhx","zrc"]
subject=subjects[2]
path=os.path.join(data_root,subject)
input_file=os.listdir(path)[30]
input_dir=os.path.join(path,input_file)
print(input_dir)
df=pd.read_csv(input_dir,header=None).T
df=df.iloc[360:600].reset_index(drop=True)
n_channel=df.shape[1]
df.columns=['CH{}'.format(i+1) for i in range(n_channel)]
df=df.rolling(window=5).mean()
fig,ax1=plt.subplots(figsize=(7,3))
df.plot(ax=ax1,legend=False)
ax1.set_xlim(0,df.shape[0])
ax1.set_xlabel('Time/ms')
ax1.set_ylabel('Voltage/mV')
plt.savefig(os.path.join("./figure","signal_sample_curve.pdf"),bbox_inches="tight",dpi=300)
#Ledamap
# sensor_loc_list=np.array([3,11,2,12,10,9,15,16,8,5,7,1,4,6,13,14])-1
# group_list=[sensor_loc_list[:6],sensor_loc_list[6:12],sensor_loc_list[12:]]
# band=["I","II","III"]
# fig2,axes=plt.subplots(1,3,figsize=(12,5))
# grid = plt.GridSpec(1, 3, wspace=0.6,hspace=0.4)
# for i in range(3):
# band_data=df.iloc[:,group_list[i]]
# # print(band_data.info())
# ax=plt.subplot(grid[0,i],polar=True)
# values=band_data.mean()
# angles=np.linspace(0,2*np.pi,len(values),endpoint=False)
# values=np.concatenate((values,[values[0]]))
# angles=np.concatenate((angles,[angles[0]]))
# ax.plot(angles,values,'o-',color='#ff0000',linewidth=2)
# ax.fill(angles,values,color='#ff0000',alpha=0.25)
# ax.set_thetagrids(angles*180/np.pi,band_data.columns)
# if i!=2:
# ax.set_xlabel("\nArmband {}".format(band[i]),labelpad=10)
# else:
# ax.set_xlabel("Armband {}".format(band[i]),labelpad=10)
# # plt.suptitle(input_dir)
# plt.savefig(os.path.join("./figure","signal_sample_leda.pdf"),bbox_inches="tight",dpi=300)
def model_performance():
result=[]
for model_name in ["LDA","LSTM","ANN","CNN","GCN"]:
print(model_name)
args=build_args(model_name)
file_dir=os.path.join(args.output_root,"model_performance")
cm_path=file_dir+"/{}_CM.npy".format(model_name)
acc_path=file_dir+"/{}_ACC.npy".format(model_name)
cm=np.load(cm_path)
acc=np.load(acc_path)
print(acc)
result.append(acc)
CM=np.rint(cm.mean(axis=0)) #rint取整
print(CM)
utils.v_confusion_matrix(CM,args.part_actions,save_path="./figure"+"/{}_CM.pdf".format(args.model_name))
result=pd.DataFrame(result,index=["LDA","LSTM","ANN","CNN","GCN"],columns=["S1","S2","S3","S4","S5"])
result["mean"]=result.mean(axis=1)
result["std"]=result.std(axis=1)
print(result)
result.to_csv(file_dir+"/model_performance.csv")
def used_time():
result=[]
for model_name in ["LDA","LSTM","ANN","CNN","GCN"]:
print(model_name)
args=build_args(model_name)
file_dir=os.path.join(args.output_root,"model_usedtime")
time_path=file_dir+"/{}_Time.npy".format(model_name)
usedtime=np.load(time_path)
result.append(usedtime)
result=pd.DataFrame(result,index=["LDA","LSTM","ANN","CNN","GCN"],columns=["S1","S2","S3","S4","S5"])
result["mean"]=result.mean(axis=1)
result["std"]=result.std(axis=1)
print(result)
result.to_csv(file_dir+"/model_usedtime.csv")
def point_optimalzation_analysis():
sensor_loc_list=np.array([3,11,2,12,10,9,15,16,8,5,7,1,4,6,13,14])-1
loc_list=["B1","B2","B3","B4","B5","B6","C1","C2","C3","C4","C5","C6","A1","A2","A3","A4"]
loc_dict=dict(zip(sensor_loc_list,loc_list))
print(loc_dict)
args=build_args(dataset="dataset2")
save_path=os.path.join(args.output_root,"channel_reduce")
sub_index=0
model_list=["LSTM","ANN","CNN","GCN"]
optimal_result=pd.DataFrame(index=model_list,columns=list(range(1,17))[::-1])
fig,ax=plt.subplots(figsize=(8,4))
for index,model_name in enumerate(model_list):
with open(os.path.join(save_path,'channel_reduction_result_{}_{}_s{}.json'.format(args.dataset,model_name,sub_index)),"r") as json_file:
result_dict=json.load(json_file)
ch_list=[]
score_list=[]
keys=list(result_dict.keys())
for n,key in enumerate(keys):
if n == len(keys)-1:
_keys=list(result_dict[key].keys())
if key == _keys[0]:
ch_list.append(_keys[0])
ch_list.append(_keys[1])
score_list.append(result_dict[key][_keys[0]])
score_list.append(result_dict[key][_keys[1]])
else:
ch_list.append(_keys[1])
ch_list.append(_keys[0])
score_list.append(result_dict[key][_keys[1]])
score_list.append(result_dict[key][_keys[0]])
else:
ch_list.append(key)
score_list.append(result_dict[key][key])
# print(ch_list)
opti_loc=[loc_dict[int(ch)] for ch in ch_list]
optimal_result.iloc[index,:]=opti_loc
# print(score_list)
if model_name == "GCN":
model_name="GAM-Net(Ours)"
plt.plot(score_list[::-1],marker="o",markersize=5,label=model_name)
print(optimal_result)
optimal_result.to_csv(os.path.join(save_path,"point_optimization_result_{}.csv".format(args.dataset)))
plt.legend(loc=4)
plt.xticks(range(len(score_list)),list(range(1,17)))
plt.xlabel("Sensor Number k")
plt.ylabel("Accuracy")
plt.savefig(os.path.join("./figure","point_optimization_result_{}.pdf".format(args.dataset)),dpi=400,bbox_inches='tight')
def ablation_study_optim():
args=build_args(dataset="dataset2")
save_path=os.path.join(args.output_root,"channel_reduce")
with open(os.path.join(save_path,'channel_reduction_result_dataset2_GCN_s0.json'),"r") as json_file:
result_dict=json.load(json_file)
ch_list=[]
score_list=[]
keys=list(result_dict.keys())
for n,key in enumerate(keys):
if n == len(keys)-1:
_keys=list(result_dict[key].keys())
if key == _keys[0]:
ch_list.append(_keys[0])
ch_list.append(_keys[1])
score_list.append(result_dict[key][_keys[0]])
score_list.append(result_dict[key][_keys[1]])
else:
ch_list.append(_keys[1])
ch_list.append(_keys[0])
score_list.append(result_dict[key][_keys[1]])
score_list.append(result_dict[key][_keys[0]])
else:
ch_list.append(key)
score_list.append(result_dict[key][key])
print(ch_list)
print(score_list)
with open(os.path.join(save_path,'channel_reduction_result_random_select.json'),"r") as json_file:
result_dict=json.load(json_file)
random_score_list=[]
for key in list(result_dict.keys()):
random_score_list.append(np.mean(list(result_dict[key].values())))
random_score_list.append(0.984)
fig,ax=plt.subplots(figsize=(8,4))
plt.plot(score_list[::-1],marker="o",color="#CC0000",markersize=5,label="Ours")
plt.plot(random_score_list,linestyle="--",color="gray",marker="o",markersize=5,label="Random")
plt.legend(loc=4)
plt.xticks(range(len(score_list)),list(range(1,17)))
plt.xlabel("Sensor Number k")
plt.ylabel("Accuracy")
plt.savefig(os.path.join("./figure","ablation_study_random_select.pdf"),dpi=400,bbox_inches='tight')
if __name__=="__main__":
temporal_curve()
model_performance()
used_time()
point_optimalzation_analysis()
ablation_study_optim()
# plt.show()