通过加权平均来平滑(smooth)一维数据

import matplotlib.pyplot as plt
import numpy as np
import copy

def average_smooth(arr, weights=[1, 2, 3, 5, 3, 2, 1]):
    result = copy.copy(arr)
    
    #
    # 开始索引：len(weights) // 2
    # 结束索引：len(arr) - len(weights)
    # 其实就是掐头去尾进行加权平均
    #

    for i in range(len(weights) // 2, len(arr) - len(weights)):
        result[i] = np.average(result[i:i+len(weights)], axis=0, weights=weights)

    #
    # np.average(result[i:i+len(weights)], axis=0, weights=weights)
    # 以上调用的展开其实就是：
    # total_weight = np.sum(weights)
    # np.sum([arr[i + wi] * weights[wi] for wi in range(len(weights))]) / total_weight
    #
        
    return result

if __name__ == '__main__':

    # 原始数据
    original_data = np.random.random(100)

    # 加权平均来平滑数据，消除锯齿
    smoothed_data = average_smooth(original_data)

    plt.figure()
    plt.grid()
    plt.plot(original_data)
    plt.plot(smoothed_data, 'g')
    plt.show()

蓝色为原始数据，绿色是平滑过后的数据