针对无线信号辐射源识别的要求，已有十五个辐射源的数据，分别存储在十五个.mat文件中，数据表现为a+bi的复数形式。要求对每一个文件中的数据分别取十个样本点计算其聚类中心，并以得到的十五个聚类中心点为聚类中心对所有文件中的所有剩余数据一起进行密度峰值聚类，请使用python实现并给出代码

以下是 Python 代码实现：

import numpy as np
from sklearn.cluster import KMeans
from sklearn.neighbors import NearestNeighbors

# 读取.mat文件中的数据
data = []
for i in range(1, 16):
    filename = 'data{}.mat'.format(i)
    mat = scipy.io.loadmat(filename)
    data.append(mat['a'] + 1j * mat['b'])

# 对每个文件中的数据取十个样本点计算聚类中心
centers = []
for i in range(15):
    kmeans = KMeans(n_clusters=1, init='k-means++', n_init=10, max_iter=300, random_state=)
    kmeans.fit(data[i][:10].real)
    centers.append(kmeans.cluster_centers_[] + 1j * kmeans.cluster_centers_[])

# 密度峰值聚类
X = np.concatenate(data)
nbrs = NearestNeighbors(n_neighbors=10).fit(X.real)
distances, indices = nbrs.kneighbors(X.real)
rho = np.zeros(X.shape[])
for i in range(X.shape[]):
    rho[i] = 1 / np.mean(distances[i])
delta = np.zeros(X.shape[])
for i in range(X.shape[]):
    delta[i] = np.min(distances[indices[i][1:]])
gamma = rho * delta
order = np.argsort(-gamma)
cluster_centers = []
for i in range(X.shape[]):
    if gamma[order[i]] < .01:
        break
    is_center = True
    for j in range(len(cluster_centers)):
        if np.abs(X[order[i]] - cluster_centers[j]) < delta[order[i]]:
            is_center = False
            break
    if is_center:
        cluster_centers.append(X[order[i]])

# 输出聚类中心
for i in range(15):
    print('File {}:'.format(i+1))
    for j in range(len(cluster_centers)):
        distance = np.abs(cluster_centers[j] - centers[i])
        print('Cluster center {}: distance = {}'.format(j+1, distance))