线性回归实践

官方文档：http://scikit-learn.org/stable/modules/linear_model.html

一、线性回归实践

1、导入相关库，并查看数据情况

 

 2、对于预测的变量，查看分布情况

3、对于几个特征，查看与因变量的关系

结论：三个特征，前两个与销量呈现明显的线性关系，第三个关系比较弱

4、建立模型，做预测

 

 　　

以下是所有的代码，这样比较好复制出来

#导入库
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
#显示中文
import matplotlib as mpl
mpl.rcParams['font.sans-serif'] = [u'SimHei']
mpl.rcParams['axes.unicode_minus'] = False
#读取数据
data = pd.read_csv('8.Advertising.csv',index_col=0)
data.head()
fig,ax = plt.subplots()
ax.hist(data['Sales'])
fig.set_size_inches(10,6)
plt.title('销量分布直方图',size=20)   
fig,ax = plt.subplots(3,1)
fig.set_size_inches(8,14)
ax[0].scatter(data['TV'],data['Sales'])
ax[0].set_xlabel('TV&Sales',size=16)
ax[1].scatter(data['Radio'],data['Sales'],color='r')
ax[1].set_xlabel('Radio&Sales',size=16)
ax[2].scatter(data['Newspaper'],data['Sales'],color='y')
ax[2].set_xlabel('Newspaper&Sales',size=16)
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
#数据切分，训练集，测试集
x = data[['TV','Radio','Newspaper']]
y = data['Sales']
x_train,x_test,y_train,y_test = train_test_split(x,y,random_state=1)
#训练模型
linreg = LinearRegression()
linreg.fit(x_train,y_train)
#模型参数
linreg.coef_,linreg.intercept_
#预测结果，并查看误差值
y_hat = linreg.predict(x_test)
mean_squared_error(y_hat,y_test)
#用图形显示差异
fig,ax = plt.subplots()
ax.plot(np.arange(len(x_test)),y_test)
ax.plot(np.arange(len(x_test)),y_hat)
fig.set_size_inches(12,6)
plt.legend(['y_test','y_hat'],fontsize=12)
from sklearn.linear_model import Ridge
from sklearn.linear_model import Lasso
from sklearn.model_selection import GridSearchCV
#lasso 回归
lasso = Lasso()
lasso_model = GridSearchCV(lasso,param_grid=({'alpha':[1,2,3,4,5]}),cv=10)  #使用网格搜索确定lass回归的参数alpha，10折交叉验证
lasso_model.fit(x_train,y_train)
lasso_model.best_params_   #输出最佳参数
#预测测试集，并输出误差
y_hat_lasso = lasso_model.predict(x_test)
print(mean_squared_error(y_test,y_hat_lasso))
#ridge回归
ridge = Ridge()
ridge_model = GridSearchCV(ridge,param_grid=({'alpha':np.linspace(0,5,10)}),cv=10)  #从0到5等差数列一共十个数,等比数列np.logspace
ridge_model.fit(x_train,y_train)
print(ridge_model.best_params_)
#预测测试集，并输出误差
y_hat_ridge = ridge_model.predict(x_test)
print(mean_squared_error(y_test,y_hat_ridge))

二、Logistic回归实践

 

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
#显示中文
import matplotlib as mpl
mpl.rcParams['font.sans-serif'] = [u'SimHei']
mpl.rcParams['axes.unicode_minus'] = False
#读取鸢尾花数据
from sklearn.datasets import load_iris
iris = load_iris()
x = iris.data
y = iris.target
fig,ax = plt.subplots(1,4)
fig.set_size_inches(18,4)
for i in range(4):
    ax[i].hist(data[:,i])
    ax[i].set_xlabel(iris.feature_names[i],fontsize=12)
data.mean(axis=0)
fig,ax = plt.subplots()
fig.set_size_inches(10,5)
ax.boxplot(x,labels=iris.feature_names)
#建立模型
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
#切分数据，并训练模型
x_train,x_test,y_train,y_test = train_test_split(x,y,random_state=1)
lr = LogisticRegression()
lr.fit(x_train,y_train)
#预测数据，输出准确率
y_hat = lr.predict(x_test)
print(accuracy_score(y_hat,y_test))