andrews_curve_iris.py

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.datasets import load_iris

def andrews_curve(data, weights=None):
    num_variables = data.shape[1]
    t = np.linspace(0, 2*np.pi, 100)
    curve = np.zeros((len(t), 2))

    for i in range(num_variables):
        if weights is None:
            weights = np.ones(num_variables)
        theta = t * i
        x = np.cos(theta) * weights[i]
        y = np.sin(theta) * weights[i]
        curve[:, 0] += x
        curve[:, 1] += y

    curve[:, 0] /= np.sum(weights)
    curve[:, 1] /= np.sum(weights)

    return curve

# 붓꽃 데이터셋 로드
iris = load_iris()
data = iris.data
target = iris.target
feature_names = iris.feature_names

# 데이터 프레임 생성
df = pd.DataFrame(data, columns=feature_names)

# 앤드루스 곡선 그리기
curve = andrews_curve(df)
colors = ['r', 'g', 'b']
for i in range(len(curve)):
    plt.plot(curve[i, 0], curve[i, 1], color=colors[target[i]], marker='o')
plt.title("Andrews Curve - Iris Dataset")
plt.show()