github地址

import pandas as pdimport numpy as npimport matplotlib.pyplot as plt%matplotlib inline复制代码

/anaconda3/envs/py35/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88  return f(*args, **kwds)复制代码

from sklearn.linear_model import LinearRegressionfrom sklearn.model_selection import train_test_splitfrom sklearn.metrics import mean_squared_error复制代码

plt.style.use('ggplot')复制代码

data = pd.read_csv('Advertising.csv')复制代码

data.head()复制代码

plt.scatter(data.TV, data.sales)复制代码

    
     复制代码
    

plt.scatter(data.radio, data.sales)复制代码

    
     复制代码
    

plt.scatter(data.newspaper, data.sales)复制代码

    
     复制代码
    

x = data[['TV', 'radio', 'newspaper']]复制代码

y = data.sales复制代码

x_train, x_test, y_train, y_test = train_test_split(x, y)复制代码

len(x_train), len(y_train)复制代码

(150, 150)复制代码

len(x_test)复制代码

50复制代码

model = LinearRegression()复制代码

model.fit(x_train, y_train)复制代码

LinearRegression(copy_X=True, fit_intercept=True, n_jobs=1, normalize=False)复制代码

model.coef_复制代码

array([ 0.04466416,  0.19594144, -0.00469486])复制代码

for i in zip(x_train.columns, model.coef_):    print(i)复制代码

('TV', 0.04466415613441986)('radio', 0.1959414384329583)('newspaper', -0.0046948632484331895)复制代码

mean_squared_error(model.predict(x_test), y_test)复制代码

3.927556655626268复制代码