from copy import deepcopy
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt

plt.rcParams['figure.figsize'] = (16, 9)
plt.style.use('ggplot')

# 创建示例数据
def create_sample_data():
    np.random.seed(42)
    cluster1 = np.random.normal(loc=[0, 0], scale=1, size=(100, 2))
    cluster2 = np.random.normal(loc=[10, 5], scale=1.5, size=(100, 2))
    cluster3 = np.random.normal(loc=[5, 10], scale=1.2, size=(100, 2))
    data = np.vstack([cluster1, cluster2, cluster3])
    df = pd.DataFrame(data, columns=['V1', 'V2'])
    df.to_csv('xclara.csv', index=False)

# 如果文件不存在则创建
try:
    data = pd.read_csv('xclara.csv')
except:
    create_sample_data()
    data = pd.read_csv('xclara.csv')

f1 = data['V1'].values
f2 = data['V2'].values
X = np.array(list(zip(f1, f2)))

# 距离计算函数
def dist(a, b, ax=1):
    return np.linalg.norm(a - b, axis=ax)

# 设置聚类数(cluster)
k = 3

# 随机初始化质心
C_x = np.random.uniform(np.min(f1), np.max(f1), size=k)
C_y = np.random.uniform(np.min(f2), np.max(f2), size=k)
C = np.array(list(zip(C_x, C_y)), dtype=np.float32)

# 绘制初始数据点
plt.scatter(f1, f2, c='black', s=7)
plt.scatter(C_x, C_y, marker='*', s=200, c='red')
plt.title("Initial Data Points and Centroids")
plt.show()

# 初始化变量
C_old = np.zeros(C.shape)
clusters = np.zeros(len(X))
iteration_flag = dist(C, C_old, None)  # 初始距离
tmp = 1

# K-Means循环
while iteration_flag != 0 and tmp < 20:
    # 1. 分配点到最近的质心，划到簇里
    for i in range(len(X)):
        distances = dist(X[i], C, 1)  # 计算点到所有质心的距离
        cluster_idx = np.argmin(distances)  # 找到最近的质心索引
        clusters[i] = cluster_idx
    
    # 2. 保存旧质心
    C_old = deepcopy(C)
    
    # 3. 更新质心位置
    for i in range(k):
        # 获取属于当前簇的所有点
        points = X[clusters == i]
        if len(points) > 0:
            C[i] = np.mean(points, axis=0)
        else:
            # 如果簇为空，重新初始化质心
            C[i] = np.random.uniform(np.min(X), np.max(X), size=2)
    
    print(f'Iteration {tmp}')
    tmp += 1
    
    # 4. 计算质心移动距离
    iteration_flag = dist(C, C_old, None)
    print(f'Centroid movement distance: {iteration_flag:.4f}')

# 绘制最终聚类结果
colors = ['r', 'g', 'b', 'y', 'c', 'm']
fig, ax = plt.subplots()
for i in range(k):
    points = X[clusters == i]  # 选择属于当前簇的点
    ax.scatter(points[:, 0], points[:, 1], s=7, c=colors[i])
ax.scatter(C[:, 0], C[:, 1], marker="*", s=200, c='black')
plt.title("Final Clustering Result")
plt.show()