核密度估计求带宽

# -*- coding: UTF-8 -*-
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.neighbors import KernelDensity
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import LeaveOneOut
sns.set()


def make_data(N, f=0.3, rseed=1):
    rand = np.random.RandomState(rseed)
    x = rand.randn(N)
    x[int(f * N):] += 5  # f * N以后全部增加5
    return x


x = make_data(20)
x_d = np.linspace(-4, 8, 1000)
# 初始化并埋入数据
kde = KernelDensity(bandwidth=1.0, kernel='gaussian')
kde.fit(x[:, None])
# score_samples 返回概率密度的对数值

logprob = kde.score_samples(x_d[:, None])
plt.fill_between(x_d, np.exp(logprob), alpha=0.5)
plt.plot(x, np.full_like(x, -0.01), '|k', markeredgewidth=1)
plt.ylim(-0.02, 0.22)

bandwidths = 10 ** np.linspace(-1, 1, 100)
grid = GridSearchCV(KernelDensity(kernel='gaussian'), {'bandwidth': bandwidths}, cv=LeaveOneOut())
grid.fit(x[:, None])
print('grid.best_params_ is: ', grid.best_params_)
plt.show()