Pandas⑫ (回帰直線)

回帰直線

前回記事の続きとしまして、今回は 回帰直線 を描画してみます。

(CSVファイルの読み込みと回帰モデルは前回記事で実行したものを利用します。)

散布図 を描画して、その上に 回帰直線 を描きます。

[Google Colaboratory]

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
from matplotlib import pyplot as plt
%matplotlib inline

xx = np.arange(20, 40) # 20~40の等差数列を生成
yy = model.predict(xx[:, np.newaxis]) # 回帰分析結果でxxに対するy値を予測する

plt.plot(xx, yy, label='predicted') # 回帰直線をプロット
plt.plot(x, y, 'o', label='sales') # x、yの散布図をプロット

plt.xlabel('temp') # x軸のラベル
plt.ylabel('sale') # y軸のラベル

plt.xlim(20, 40) # x軸の範囲を設定
plt.ylim(0, 500) # y軸の範囲を設定

plt.legend() # 凡例を表示

[実行結果]

回帰直線 を表示することができました。

実測データの最高気温の最小値は24℃、最大値は35℃であり、この間では気温が 1℃上昇 すると、計算上 33.7408(回帰係数)ずつ売上が増えることを示しています。