-
Notifications
You must be signed in to change notification settings - Fork 2
/
PredictPath.py
262 lines (227 loc) · 9.52 KB
/
PredictPath.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
"""
Predict Path
Usage:
PredictPath.py [--pre] [--cor]
PredictPath.py (-h | --help)
Options:
-h --help Show this screen
--pre Predict Path
--cor Generate Correlation
"""
import pandas as pd
import numpy as np
import math
import matplotlib.pyplot as plt
import sys
from sklearn import linear_model
from docopt import docopt
from sklearn.preprocessing import PolynomialFeatures
from utils.frechetdist import frdist
# 计算角度
def calc_angle(x_point_s,y_point_s,x_point_e,y_point_e):
angle = 0
y_se = y_point_e - y_point_s
x_se = x_point_e - x_point_s
if x_se == 0 and y_se > 0:
angle = 360
if x_se == 0 and y_se < 0:
angle = 180
if y_se == 0 and x_se > 0:
angle = 90
if y_se == 0 and x_se < 0:
angle = 270
if x_se > 0 and y_se > 0:
angle = math.atan(x_se / y_se) * 180 / math.pi
elif x_se < 0 and y_se > 0:
angle = 360 + math.atan(x_se / y_se) * 180 / math.pi
elif x_se < 0 and y_se < 0:
angle = 180 + math.atan(x_se / y_se) * 180 / math.pi
elif x_se > 0 and y_se < 0:
angle = 180 + math.atan(x_se / y_se) * 180 / math.pi
return angle
if __name__ == '__main__':
arguments = docopt(__doc__)
# 设置最大递归限制
sys.setrecursionlimit(15000)
# 数据集路径
datasets_path = './data/Path/situation_0901.xlsx'
# 保存路径
save_path = './out/'
# 设置图像为1600x900
plt.figure(figsize=(19.2, 10.8))
# 计算数据集Sheet标签
file = pd.ExcelFile(datasets_path)
sheets_name = file.sheet_names
# 构建完整数据集
df = pd.read_excel(datasets_path, sheet_name=sheets_name)
dataset = []
lon_lat_data_y = []
for v in df:
data_item = df[v]
data_item = data_item[data_item.keys().to_numpy()].to_numpy()
# 经纬度时序数据
lon_lat_data_y_item = data_item[:, 9:11]
dataset.append(data_item)
lon_lat_data_y.append(np.array(lon_lat_data_y_item, dtype=np.float))
dataset = np.array(dataset, dtype=np.object)
lon_lat_data_y = np.array(lon_lat_data_y, dtype=np.object)
lon_lat_data_x = []
# 构建时序标签
for i in range(lon_lat_data_y.shape[0]):
data_len = lon_lat_data_y[i].shape[0]
data_y = np.arange(0, data_len)
lon_lat_data_x.append(data_y)
lon_lat_data_x = np.array(lon_lat_data_x, dtype=np.object)
if arguments["--pre"]:
# 表格纵轴
pred_y_index = []
pred_data = []
pred_step = 7000
for i in range(lon_lat_data_y.shape[0]):
lon_lat_data_y_item = lon_lat_data_y[i]
lon_lat_data_x_item = lon_lat_data_x[i].reshape(-1, 1)
# 获取对象唯一编号
name = dataset[i][0][0]
# 构建线性回归模型
regModel = linear_model.LinearRegression()
# 构建输入特征
poly_reg = PolynomialFeatures(degree=2)
# 获取输入特征
lon_lat_data_x_item = poly_reg.fit_transform(lon_lat_data_x_item)
# 拟合现有数据
regModel.fit(lon_lat_data_x_item, lon_lat_data_y_item)
# 数据预测
lon_lat_index_pred = np.arange(lon_lat_data_x_item.shape[0] + 1,
lon_lat_data_x_item.shape[0] + 1 + pred_step).reshape(-1, 1)
# 预测的输入特征转化为构建的特征
lon_lat_index_pred = poly_reg.transform(lon_lat_index_pred)
lon_lat_data_pred = np.around(regModel.predict(lon_lat_index_pred), 6)
# 构建保存数据
lon_lat_data_pred = np.array(lon_lat_data_pred).flatten()
pred_data.append(lon_lat_data_pred)
pred_y_index.append(name)
print('Predict Path {0}/{1}'.format(i + 1, lon_lat_data_y.shape[0]))
# 创建数据的列标题
columns_name = []
for i in range(pred_step):
columns_name.append('longitude-{0}'.format(i + 1))
columns_name.append('latitude-{0}'.format(i + 1))
for i in range(len(pred_data)):
# 预测数据
pred_data_item = pred_data[i]
longitude_pred_data_item = pred_data_item[::2]
latitude_pred_data_item = pred_data_item[1::2]
# 历史数据
data_item = lon_lat_data_y[i]
data_item = data_item.flatten()
longitude_data_item = data_item[::2]
latitude_data_item = data_item[1::2]
longitude_data_item = longitude_data_item[-pred_step:]
latitude_data_item = latitude_data_item[-pred_step:]
# 计算预测线与正北角度
angle = calc_angle(longitude_pred_data_item[0], latitude_pred_data_item[0],
longitude_pred_data_item[-1], latitude_pred_data_item[-1])
# 清空图像
plt.clf()
plt.title('{} Predict, Angle {:.2f}'.format(pred_y_index[i], angle))
plt.plot(longitude_data_item, latitude_data_item, label='History', linestyle='--',
color='orange', marker='^', markerfacecolor='orange')
plt.plot(longitude_pred_data_item, latitude_pred_data_item, label='Predict', marker='o')
# 添加标记
plt.text(longitude_pred_data_item[0], latitude_pred_data_item[0], '1')
plt.text(longitude_pred_data_item[-1], latitude_pred_data_item[-1], len(longitude_pred_data_item))
plt.xlabel('Longitude')
plt.ylabel('Latitude')
plt.legend()
# 保存预测图像
plt.savefig(save_path + '{0}_predict.png'.format(pred_y_index[i]))
# 创建数据
df = pd.DataFrame(pred_data, index=pred_y_index, columns=columns_name)
# 写入Excel文件
df.to_excel(save_path + 'pred_data.xlsx', sheet_name='pred_data')
print('Predict Data Finish...')
if arguments["--cor"]:
# 相似性矩阵
sim_arr = []
# 表格纵轴
sim_index = []
count = 1
for i in range(len(lon_lat_data_y)):
lon_lat_data_item = lon_lat_data_y[i]
sim_arr_item = []
# 获取对象唯一编号
name = dataset[i][0][0]
sim_index.append(name)
for k in range(len(lon_lat_data_y)):
if i == k:
sim_arr_item.append(-9999)
continue
# 获得要计算相关性的对象
lon_lat_data_com = lon_lat_data_y[k]
# 计算Fréchet distance（弗雷歇距离）相似性
# sim = frdist(lon_lat_data_item, lon_lat_data_com)
# print('{0}-{1} Sim = {2}'.format(dataset[i][0][0], dataset[k][0][0], sim))
sum_sim = 0
# 数据长短不一时取短的
data_len = min(len(lon_lat_data_item), len(lon_lat_data_com))
last_dis = None
for j in range(data_len):
A = lon_lat_data_item[j]
B = lon_lat_data_com[j]
# 计算欧式距离
dis = np.linalg.norm(A - B)
if last_dis is None:
last_dis = dis
else:
# 计算距离的变化
diff = (dis - last_dis)
sum_sim = sum_sim + diff
sim = sum_sim / data_len
# 保存相似性数据
sim_arr_item.append(sim)
print('Correlation Compute {0}/{1}'.format(count, len(lon_lat_data_y) ** 2 - len(lon_lat_data_y)))
count = count + 1
# 保存相似性数据
sim_arr.append(sim_arr_item)
sim_arr = np.array(sim_arr)
sim_arr = sim_arr.flatten()
delete_index = []
for i in range(len(sheets_name)):
delete_index.append(i * (len(sheets_name) + 1))
sim_arr_tmp = np.delete(sim_arr, delete_index)
# 归一化数据
sim_min, sim_max = np.min(sim_arr_tmp), np.max(sim_arr_tmp)
# 限制数据小于1
sim_arr_tmp = (1 - (sim_arr_tmp - sim_min) / (sim_max - sim_min))
# 插入自相关的值
insert_index = []
for i in range(len(sheets_name)):
insert_index.append(i * len(sheets_name))
sim_arr = np.insert(sim_arr_tmp, insert_index, 1)
# 重新变为二维数组
sim_arr = sim_arr.reshape(-1, len(sheets_name))
# 显示相关性热力图
fig, ax = plt.subplots(1, 1, figsize=(19.2, 10.8))
im = ax.imshow(sim_arr)
plt.colorbar(im)
# 绘制坐标轴
ax.set_xticks(np.arange(len(sim_index)))
ax.set_yticks(np.arange(len(sim_index)))
ax.set_xticklabels(sim_index)
ax.set_yticklabels(sim_index)
plt.setp(ax.get_xticklabels(), rotation=45, ha="right", rotation_mode="anchor")
# 绘制热力图中的数值
for i in range(len(sheets_name)):
for j in range(len(sheets_name)):
text = ax.text(j, i, '{:.3f}'.format(sim_arr[i, j]), ha='center', va='center', color="w")
fig.tight_layout()
plt.title('Object Correlation Heatmap')
# 保存图片
plt.savefig(save_path + 'object-correlation-heatmap.png')
# 创建数据
df = pd.DataFrame(sim_arr, index=sim_index, columns=sim_index)
# 写入Excel文件
df.to_excel(save_path + 'sim_data.xlsx', sheet_name='sim_data')
print('Calculate Similarity Finish...')
# 显示所有图表
plt.show()