THQA/vvmf.py at main · zyj-2000/THQA · GitHub

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
import pandas as pd
import math
import csv
import numpy as np


def get_points(n):
    """
    用户输入n,输出从y=e^(-x)函数中的n个(x,y)
    使得相邻两个x之间的距离相等,且所有y值相加和为1
    """
    # 定义y=e^(-x)函数
    def f(x):
        return math.exp(-1*x)

    # 计算所有y值之和
    total = 0
    for i in range(n):
        total += f(i)

    # 计算x的起始值和终止值
    x_start = 0
    x_end = n - 1

    # 计算相邻x之间的距离
    dx = (x_end - x_start) / (n - 1)

    # 初始化结果列表
    points = []

    # 遍历x值,计算相应的y值
    x = x_start
    for i in range(n):
        y = f(x) / total
        points.append((x, y))
        x += dx

    return points

def csv_write(subject, predict):
    with open('./vvmf_thzs_a1.csv', 'a+', newline='') as csvfile:
        writer = csv.writer(csvfile)
        predict = predict.tolist()
        predict.insert(0, subject)  # 在predict数组的第一个位置插入subject
        rows = [predict]
        print(rows)
        writer.writerows(rows)

def process(previous_image,png_list,predict_list):
    num = len(png_list)
    scores = []
    for jj in range(len(predict_list)):
        if num == 0:
            pass
        else:
            score = 0
            points = get_points(num)
            weights = []
            for i in range(num):
                weights.append(points[i][1])
            for j in range(num):
                score = score + predict_list[j] * weights[num-1-j]
            scores.append(score)
        # print(previous_image,scores)
        # csv_write(previous_image,score)

def process_g(previous_image,png_list,predict_list):
    num = len(png_list)
    scores = []

    if num == 0:
        pass
    else:
        score = np.zeros(len(predict_list[0]))
        points = get_points(num)
        weights = []
        for i in range(num):
            weights.append(points[i][1])
        for j in range(num):
            # print(predict_list[j])
            score = score + np.array(predict_list[j]) * weights[num-1-j]
           # print(score.shape)
        scores.append(score.tolist())
    # print(previous_image,len(scores))
        csv_write(previous_image,score)


# 读取test.csv文件
df = pd.read_csv('lib_c.csv')

# 初始化变量
previous_image = None
png_list = []
predict_list = []
# 逐行读取并处理数据
for index, row in df.iterrows():
    current_image = row['Image']
    current_png = row['png']
    # current_predict = row[2:28]
    current_predict = np.array(row[2:28]).tolist() # 将current_predict转换为列表形式

    # 判断Image列的内容是否和前一个不一样
    if current_image != previous_image:
        print("----------------------------")
        process_g(previous_image,png_list,predict_list)
        # # 如果和前一个不一样，新建png_list和predict_list
        # # print(png_list,predict_list)
        png_list = []
        predict_list = []
        png_list.append(current_png)
        predict_list.append(current_predict)
        previous_image = current_image
    else:
        # 如果和前一个一样，继续将该行的png和predict放入对应数组中
        png_list.append(current_png)
        predict_list.append(current_predict)
        previous_image = current_image
process_g(previous_image,png_list,predict_list)