线性回归简介

Y = a + bX

1import pandas as pd
2import numpy as np
3
4import matplotlib.pyplot as plt
5import seaborn as sns
6
7df_advertising = pd.read_csv('advertising.csv', index_col=0)
8print('数据集')
9print(df_advertising.head())
10
11print('\n数值型列的详细信息')
12print(df_advertising.describe())
13
14print('\n数据集中空值的数量')
15print(df_advertising.isna().sum())

1plt.xlabel('电视广告支出')
2plt.ylabel('销售额')
3sns.scatterplot(data=df_advertising, x='TV', y='sales')
4plt.savefig('output/output.png', dpi=800)

1from sklearn.linear_model import LinearRegression
2
3lr = LinearRegression()
4lr.fit(df_advertising[['TV']], df_advertising['sales'])

1predictions = lr.predict(df_advertising[['TV']])
2
3plt.xlabel('电视广告支出')
4plt.ylabel('销售额')
5sns.scatterplot(data=df_advertising, x='TV', y='sales')
6plt.plot(df_advertising['TV'], predictions, color='r')
7plt.savefig('output/output.png', dpi=800)