-
Notifications
You must be signed in to change notification settings - Fork 0
/
medical_cost_predict.py
124 lines (118 loc) · 4.06 KB
/
medical_cost_predict.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
import numpy as np
import pandas as pd
import os
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
import seaborn as sns
from sklearn import metrics
from sklearn.preprocessing import LabelEncoder
from sklearn.preprocessing import PolynomialFeatures
data_file = "../统计学/insurance.csv"
data_frame = pd.read_csv(data_file) #读取数据集
#data_frame.head() #观察读取的数据是否正确
#print(data_frame.head())
#data_frame.describe() #给出各项数据的统计信息
#print(data_frame.describe())
#############对分类型数据如sex smoker region进行编码##########
#smoker
LabelEncoder_smoker = LabelEncoder()
data_frame.smoker = LabelEncoder_smoker.fit_transform(data_frame.smoker)
#sex
LabelEncoder_sex = LabelEncoder()
data_frame.sex = LabelEncoder_sex.fit_transform(data_frame.sex)
#region
LabelEncoder_region = LabelEncoder()
data_frame.region = LabelEncoder_region.fit_transform(data_frame.region)
#print(data_frame.head())
#按照与费用的相关性对各个要素进行排列
#print(data_frame.corr()['charges'].sort_values())
#用直方图表示平均有多少患者在治疗上花费了多少钱
'''
plt.hist(data_frame.charges,bins=10,alpha=0.5,histtype='bar',ec='black')
plt.title("Frequency Distribution of the charges")
plt.xlabel('Charges')
plt.ylabel('Frequency')
plt.show()
'''
raw_data_frame = pd.read_csv(data_file)
#不同地区对医疗花费的影响
'''
sns.boxenplot(x=raw_data_frame.region,y=raw_data_frame.charges,data=raw_data_frame)
plt.title("Medical charges per region")
plt.show()
'''
#吸烟与否对医疗花费的影响
'''
sns.boxplot(x=raw_data_frame.smoker,y=raw_data_frame.charges,data=raw_data_frame)
plt.title("Medical charges or Smokers and Non-Smokers")
plt.show()
'''
#进一步研究吸烟与否对医疗花费的影响
'''
f = plt.figure(figsize=(12,5))#设定plot的宽高
ax = f.add_subplot(121)
sns.distplot(data_frame[data_frame.smoker==1]['charges'],color='r',ax=ax)
ax.set_title('Medical charges for the smokers')
ax = f.add_subplot(122)
sns.distplot(data_frame[data_frame.smoker==0]['charges'],color='g',ax=ax)
ax.set_title('Medical charges for the non-smokers')
#plt.show()
'''
#性别对医疗花费的影响
'''
sns.boxplot(x=raw_data_frame.sex,y=raw_data_frame.charges,data=raw_data_frame)
plt.title("Charges by Gender")
plt.show()
'''
#探究吸烟者与不吸烟者在不同年龄段在医疗花费上的投入
'''
plt.subplot(1,2,1)
sns.distplot(data_frame[data_frame.smoker==1]['age'],color='red')
plt.title("Distribution of Smokers")
plt.subplot(1,2,2)
sns.distplot(data_frame[data_frame.smoker==0]['age'],color='blue')
plt.title("Distribution of Non-Smokers")
plt.show()
'''
#分析bmi与吸烟与否对医疗花费的影响
'''
sns.lmplot(x="bmi",y='charges',hue='smoker',data=raw_data_frame)
plt.show()
'''
#可视化回归模型
'''
sns.lmplot(x='age',y='charges',hue='smoker',data=raw_data_frame,palette='inferno_r')
plt.show()
'''
#建立回归模型
X = data_frame.iloc[:,:6].values
Y = data_frame.iloc[:,6].values
#这里用到多项式回归模型
poly_reg = PolynomialFeatures(degree=2)
X_poly = poly_reg.fit_transform(X)
#分割数据集为训练集和测试集
X_train,X_test,Y_train,Y_test = train_test_split(X_poly,Y,test_size=0.25)
#训练模型
lin_reg = LinearRegression()
lin_reg = lin_reg.fit(X_train,Y_train)
#预测准确率
#print('准确率为：',lin_reg.score(X_test,Y_test))
'''
X_poly_df = pd.DataFrame(X_poly,columns=poly_reg.get_feature_names())
print(X_poly_df.head())
X_poly_df.to_csv('coef.csv')
'''
#plt.scatter(X_test,Y_test,color='blue')
#plt.plot(X,lin_reg.predict(X_poly),color='red')
Y_predict = lin_reg.predict(X_test)
plt.figure()
plt.plot(range(len(Y_predict)),Y_predict,'r',label='charges_predict')
plt.plot(range(len(Y_predict)),Y_test,'b',label='charges_test')
plt.legend(loc='upper right')
plt.title('Polynomial Regression')
plt.show()
'''
print('coef: ',lin_reg.coef_)
print('intercept: ',lin_reg.intercept_)
'''