-
Notifications
You must be signed in to change notification settings - Fork 0
/
evaluation.py
154 lines (131 loc) · 5.95 KB
/
evaluation.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
import numpy as np
import cv2
import torch
import pandas as pd
import matplotlib.pyplot as plt
import os
import math
import arg_config
def evaluate_typhoon(predictions, targets):
# lat_long_data = pd.read_csv('result.csv')
# data = lat_long_data.loc[lat_long_data.Id == 201801, ['LAT', 'LONG', 'UP', 'DOWN', 'LEFT', 'RIGHT']].values
img_size = arg_config.img_size
predictions = predictions * img_size
targets = targets.astype(np.float64) * img_size
# pd.dataFrame()
# 轴向为行，删除索引为0的行
# predictions = np.delete(predictions, 0, axis=0)
# targets = np.delete(targets, 0, axis=0)
# target_lats = []
# target_longs = []
# pre_lats = []
# pre_longs = []
# for i in range(predictions.shape[0]):
# pre_long = 0
# pre_lat = 0
# print(predictions[i][0], targets[i][0], data)
# if predictions[i][0] < targets[i][0]:
# dist_long = (data[i][1] - data[i][4]) / 64 * (targets[i][0] - predictions[i][0])
# pre_long = data[i][1] - dist_long
# else:
# dist_long = (data[i][5] - data[i][1]) / 64 * (predictions[i][0] - targets[i][0])
# pre_long = data[i][1] + dist_long
# if predictions[i][1] < targets[i][1]:
# dist_lat = (data[i][0] - data[i][2]) / 64 * (targets[i][1] - predictions[i][1])
# pre_lat = data[i][0] - dist_lat
# else:
# dist_lat = (data[i][3] - data[i][0]) / 64 * (predictions[i][1] - targets[i][1])
# pre_lat = data[i][0] + dist_lat
# target_lats.append(data[i][0])
# target_longs.append(data[i][1])
# pre_lats.append(pre_lat)
# pre_longs.append(pre_long)
# all_overall = 0.0
# # 2.散点图,只是用用scat函数来调用即可
# for i in range(len(pre_lats)):
# print("============\n")
# print("真实纬度：", target_lats[i], "真实经度：", target_longs[i], "\n")
# print("检测纬度：", pre_lats[i], "检测经度：", pre_longs[i], "\n")
# plt.figure(num=1, figsize=(10, 10))
# myfig = plt.gcf() # Get the current figure. If no current figure exists, a new one is created using figure().
# plt.scatter(pre_longs, pre_lats, s=100, c='r')
# plt.scatter(target_longs, target_lats, s=100, c='none', marker='o', linewidths=2, edgecolors='b')
# 设置刻度范围和字体大小
# plt.ylim(10, 40)
# plt.xlim(130, 160)
# plt.yticks(rotation=0, size=15)
# plt.xticks(rotation=0, size=15)
# plt.title('路径图', fontdict={'weight': 'normal', 'size': 25})
# plt.xlabel('Longitude(°)', fontdict={'weight': 'normal', 'size': 18})
# plt.ylabel('Latitude(°)', fontdict={'weight': 'normal', 'size': 18})
# 设置图标
# plt.rcParams.update({'font.size': 18})
# plt.legend(['EF-LOCNet', 'Helianthus '])
# 显示所画的图
# plt.show()
# plt.show()
# myfig.savefig('JEBI.png', dpi=400) # save myfig
squared_diffs = [] # 存储每个预测值与目标值差的平方
for k in range(len(predictions)):
diff = targets[k] - predictions[k] # 计算差
diff = np.power(diff, 2) # 计算差的平方
squared_diffs.append(diff) # 将差的平方添加到列表中
# 计算所有差的平方的平均值
mean_squared_diff = np.mean(squared_diffs)
# 计算均方根误差（RMSE）
rmse = np.sqrt(mean_squared_diff)
print(f"RMSE: {round(rmse, 10)}")
return rmse
# def computeErrorOfLAtLon(pre_longs, pre_lats, target_longs, target_lats):
# all = 0
# for i in range(len(pre_lats)):
# a = (pre_longs[i] + pre_lats[i]) / 2
# b = (target_longs[i] + target_lats[i]) / 2
# c = math.fabs(a - b)
# all = all + c
# error = all / len(pre_lats)
# return error
def computeErrorOfLAtLon(pre_longs, pre_lats, target_longs, target_lats):
all = 0
for i in range(len(pre_lats)):
a = abs(pre_lats[i] - target_lats[i])
b = abs(pre_longs[i] - target_longs[i])
c = (a + b) / 2
all = all + c
error = all / len(pre_lats)
return error
def evaluate_detailed(predictions):
predictions = predictions * 512.
targets = np.load('data/TCLD/TEST_LABEL.npy').astype(np.float) * 512.
easy_points = np.load("data/TCLD/TestEasyIndex.npy")
hard_points = np.load("data/TCLD/TestHardIndex.npy")
mean_easy = 0.0
mean_hard = 0.0
mean_overall = 0.0
for i in range(len(easy_points)):
mean_easypont = np.power((targets[easy_points[i]] - predictions[easy_points[i]]), 2)
mean_easypont = np.sqrt(np.sum(mean_easypont))
mean_easy = mean_easy + mean_easypont
for m in range(len(hard_points)):
mean_hardpoint = np.power((targets[hard_points[m]] - predictions[hard_points[m]]), 2)
mean_hardpoint = np.sqrt(np.sum(mean_hardpoint))
mean_hard = mean_hard + mean_hardpoint
for k in range(len(predictions) - 1):
mean_overpoint = np.power((targets[k + 1] - predictions[k + 1]), 2)
mean_overpoint = np.sqrt(np.sum(mean_overpoint))
mean_overall = mean_overall + mean_overpoint
mean_easy = mean_easy / len(easy_points)
mean_hard = mean_hard / len(hard_points)
mean_overall = mean_overall / (len(predictions) - 1)
return mean_overall, mean_easy, mean_hard
def geodistance(lng1, lat1, lng2, lat2):
lng1, lat1, lng2, lat2 = map(math.radians, [float(lng1), float(lat1), float(lng2), float(lat2)]) # 经纬度转换成弧度
dlon = lng2 - lng1
dlat = lat2 - lat1
a = math.sin(dlat / 2) ** 2 + math.cos(lat1) * math.cos(lat2) * math.sin(dlon / 2) ** 2
distance = 2 * math.asin(math.sqrt(a)) * 6371 * 1000 # 地球平均半径，6371km
distance = round(distance / 1000, 3)
return distance
if __name__ == '__main__':
reg=np.load(arg_config.root+'runs/testing/mynet/regression.npy')
print(reg)