0
0

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 3 years have passed since last update.

[python] yeojohnson変換の変換前後の値をプロット

Last updated at Posted at 2021-01-16
python
import numpy as np
import pandas as pd

# ここの式をコピー
# https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.yeojohnson.html
def yeojohnson(lmbda, x):
    if x >= 0 and lmbda != 0:
        y = ((x + 1)**lmbda - 1) / lmbda
    
    elif x >= 0 and lmbda == 0:
        y = np.log(x + 1)
    
    elif x < 0 and lmbda != 2:
        y = -((-x + 1)**(2 - lmbda) - 1) / (2 - lmbda)
        
    elif x < 0 and lmbda == 2:
        y = -np.log(-x + 1)
    
    else:
        # 到達しないはず
        raise
    
    return y

# 変換する値
x_ary = np.arange(21)-10

# 2パターンの lmbda の組み合わせを試す
for lmbda_ary in [np.arange(10)*0.25, np.arange(10)*0.5]:
    
    # 変換して df にまとめる
    X_dic = {}
    for lmbda in lmbda_ary:
        X_dic[lmbda] = [yeojohnson(lmbda, float(x)) for x in x_ary]
    df = pd.DataFrame(X_dic, index=x_ary)
    
    # 描画
    df.plot(cmap='viridis', marker='.')
    plt.xlabel('元の値(x)')
    plt.ylabel('変換後の値(y)')
    plt.title('lmbda を変えた時の変換後の値の変化')
    plt.show()

output:
スクリーンショット 2021-01-16 午後11.22.05.png

0
0
0

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
0
0

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?