已知数据集为用X,Y = make_moons(n_samples=400,shuffle=True,noise=0.1,random_state=136)，其标签为0或1。用手写K-Means++算法进行模型构建，写出代码。并绘制聚类结果的散点图，写出代码。并分别用轮廓系数和准确率进行评估。写出代码

好的，以下是用手写K-Means++算法进行模型构建的代码，并绘制聚类结果的散点图，以及用轮廓系数和准确率进行评估的代码：

import numpy as np
import matplotlib.pyplot as plt
from sklearn.datasets import make_moons
from sklearn.metrics import silhouette_score, accuracy_score

class KMeansPlusPlus:
    def __init__(self, k, max_iters=300):
        self.k = k
        self.max_iters = max_iters
        
    def _init_centers(self, X):
        centers = [X[np.random.choice(len(X))]]
        while len(centers) < self.k:
            dists = [np.min([np.linalg.norm(x-c)**2 for c in centers]) for x in X]
            probs = dists / np.sum(dists)
            cum_probs = np.cumsum(probs)
            r = np.random.rand()
            for i, cp in enumerate(cum_probs):
                if r < cp:
                    centers.append(X[i])
                    break
        return np.array(centers)
    
    def _assign_clusters(self, X, centers):
        clusters = [[] for _ in range(self.k)]
        for x in X:
            dists = [np.linalg.norm(x-c) for c in centers]
            cluster_id = np.argmin(dists)
            clusters[cluster_id].append(x)
        return clusters
    
    def _update_centers(self, clusters):
        centers = []
        for i in range(self.k):
            center = np.mean(clusters[i], axis=0)
            centers.append(center)
        return np.array(centers)
    
    def fit(self, X):
        centers = self._init_centers(X)
        for _ in range(self.max_iters):
            old_centers = centers.copy()
            clusters = self._assign_clusters(X, centers)
            centers = self._update_centers(clusters)
            if np.allclose(centers, old_centers):
                break
        self.centers = centers
        self.clusters = clusters
        
    def predict(self, X):
        y_pred = []
        for x in X:
            dists = [np.linalg.norm(x-c) for c in self.centers]
            cluster_id = np.argmin(dists)
            y_pred.append(cluster_id)
        return np.array(y_pred)

# generate make_moons dataset
X, y = make_moons(n_samples=400, shuffle=True, noise=0.1, random_state=136)

# apply KMeans++ clustering with k=2
kmeans = KMeansPlusPlus(k=2)
kmeans.fit(X)
y_pred = kmeans.predict(X)

# calculate silhouette score
silhouette = silhouette_score(X, y_pred)
print("Silhouette Score: {:.2f}".format(silhouette))

# calculate accuracy
accuracy = accuracy_score(y, y_pred)
print("Accuracy: {:.2f}%".format(accuracy * 100))

# plot clustering result
fig, ax = plt.subplots(figsize=(8, 6))
colors = ['blue', 'orange']
for i, color in enumerate(colors):
    ax.scatter(X[y_pred == i, 0], X[y_pred == i, 1], c=color, label='Cluster {}'.format(i+1))
ax.scatter(kmeans.centers[:, 0], kmeans.centers[:, 1], marker='*', s=200, c='black', label='Centroids')
ax.set_xlabel('Feature 1')
ax.set_ylabel('Feature 2')
ax.set_title('KMeans++ Clustering Result')
ax.legend()
plt.show()

在这个代码中，我定义了一个名为`KMeansPlusPlus`的类，实现了手写K-Means++算法的初始化质心、分配簇和更新质心等过程。在生成`make_moons`数据集之后，我实例化了这个类，并调用`fit`方法进行训练，同时调用`predict`方法对数据进行预测。接着，我计算了轮廓系数和准确率，并将聚类结果绘制成散点图展示出来。