Data_Analysis/logisticRegression.py at master · Jorfun/Data_Analysis · 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
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
"""
====================================
logistic regression - 通过肺活量和立定跳远成绩来预测男女
====================================
"""

# 导入
import numpy as np
import matplotlib.pyplot as plt
import math


# 从csv文件获取所有数据
def loadData():

    #获取测试和训练数据
    f = open('logistic(2dim).csv', 'r')

    trainInput = []
    trainOutput = []
    testInput = []
    testOutput = []

    for idx,line in enumerate(f):
        temp = line.strip('\n').split(',')

        if(idx <= 2000):
            trainInput.append([int(temp[0]), int(temp[1])])
            trainOutput.append(int(temp[2]))
        else:
            testInput.append([int(temp[0]), int(temp[1])])
            testOutput.append(int(temp[2]))

    return trainInput, trainOutput, testInput, testOutput


# 对数据进行特征归一化处理
def normalize(x):
    xMin = x.min(axis=0)
    xMax = x.max(axis=0)
    range =  xMax - xMin
    # print("xMin \n", xMin)
    # print("xMax \n", xMax)
    # print("range \n", range, "\n\n")
    # range[0] = 1
    # xMin[0] = 0
    x_norm = (x - xMin) / range

    return x_norm


def sigmoid(x):
    if(type(x) is np.float64):
        return 1 / (1 + math.exp(-x))
    else:
        return 1 / (1 + np.exp(-x))


# 梯度下降
def gradDescent(trainInput, trainOutput):

    # 将python的list转换成numpy的ndarray，同时将输入的特征向量规范化
    x = np.array(trainInput)
    x_norm = normalize(x)
    xTrans = x_norm.transpose()
    y = np.array(trainOutput)
    # print("x \n", x, "\n\n")
    # print("x_norm \n", x_norm, "\n\n")
    # print("y \n", y, "\n\n")

    alpha = 20
    numIterations = 500
    m,n = x_norm.shape
    theta = np.ones(n)
    nOnes = np.ones(m)      #用来计算代价函数

    # 用于后面画图表示梯度下降的过程
    JTheta = []

    # 每次循环更新theta时，都使用所有的训练数据
    for i in range(0, numIterations):
        hypothesis = sigmoid(np.dot(x_norm, theta))
        loss = hypothesis - y
        # 计算J(theta) - 平均每个输入相对于当前theta的代价
        cost = np.sum( y * np.log(hypothesis) + (nOnes - y) * np.log(nOnes - hypothesis) ) / m * -1
        JTheta.append(cost)
        # print("hypothesis \n", hypothesis)
        # print("loss \n", loss, "\n\n")
        #print("Iteration %d | Cost: %f" % (i, cost))
        # 计算梯度
        gradient = np.dot(xTrans, loss) / m
        # 更新theta值
        theta = theta - alpha * gradient
        #print("theta", theta)

    return theta, JTheta


# 将模型用于测试数据集，观察每个输入的预测输出与真实值是否相同
def testModel(theta, testInput, testOutput):

    # 将python的list转换成numpy的ndarray，同时将输入的特征向量规范化
    testInput = np.array(testInput)
    testInputNormed = normalize(testInput)
    testOutput = np.array(testOutput)
    totalCount = 0
    errorCount = 0

    for idx, val in enumerate(testInputNormed):
        #print("type", type(val.dot(theta)))
        predictValue = sigmoid(val.dot(theta))

        if(predictValue > 0.5):
            predictValue = 1
        else:
            predictValue = 0

        if(predictValue != testOutput[idx]):
            errorCount += 1

        totalCount += 1

    print("Total test:", totalCount, "  Error:", errorCount, "  Error rate:", errorCount/totalCount)


#显示完整的nparray
# np.set_printoptions(threshold=np.nan)


#训练和测试模型，绘图
a1,a2,a3,a4 = loadData()
theta, JTheta = gradDescent(a1,a2)
print(theta, "\n\n")
testModel(theta,a3,a4)
#print(costValues)


# 画出梯度下降过程中代价函数值随迭代过程的变化
plt.plot(JTheta, linewidth=1.0)
plt.ylabel('cost function')
plt.xlabel('iteration times')
plt.show()


# 画出将模型用于测试数据时，预测值与真实值之间的偏差情况
# costValues = testModel(theta,a3,a4)
# print(costValues)
# plt.plot(costValues, 'ro')
# plt.ylabel('differential')
# plt.xlabel('test input')
# plt.show()


# 同时画出上面两个图
# plt.figure(1)
# plt.subplot(211)
# plt.plot(JTheta, linewidth=1.0)
# plt.ylabel('cost function')
# plt.xlabel('iteration times')

# plt.subplot(212)
# plt.plot(costValues, 'ro')
# plt.ylabel('differential')
# plt.xlabel('test input')
# plt.show()