Pythonでbiplot関数を改良

前回作った改良型のbiplotを目的変数で分かりやすくした関数を作りました。

関数

from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
def biplot_spv(df, y_name):
    y = df[y_name]
    x = df.drop(y_name, axis=1)
    ss = StandardScaler()
    ss.fit(x)
    sx = ss.transform(x)
    model = PCA()
    model.fit(sx)
    tx = model.transform(sx)
    evr = model.explained_variance_ratio_
    com = model.components_
    fac = []
    for i in range(len(evr)):
        fac.append(np.sqrt(evr[i])*com[i])
    fig, ax = plt.subplots()
    ax1 = ax.twinx()
    ax2 = ax.twiny()
    ax1.scatter(tx[:, 0], tx[:, 1], c=y, cmap="brg")
    xlim = [abs(min(fac[0])), abs(max(fac[0]))]
    ylim = [abs(min(fac[1])), abs(max(fac[1]))]
    ax2.set_xlim(-max(xlim), max(xlim))
    ax2.set_ylim(-max(ylim), max(ylim))
    for i in range(len(x.columns)):
        ax2.annotate("", xytext=[0, 0], xy=[fac[0][i], fac[1][i]],
                     arrowprops=dict(shrink=0, width=1, headwidth=6, 
                                    headlength=10, connectionstyle='arc3',
                                    facecolor='red', edgecolor='red'))
        ax2.text(fac[0][i], fac[1][i], x.columns[i])
    plt.show()

使用例

import pandas as pd

df = pd.read_csv("wine.csv")
biplot_spv(df, "Wine")

df = pd.read_csv("boston.csv")
biplot_spv(df, "PRICE")