Python中preprocessing.StandardScaler()在线标准化方法案例

最新推荐文章于 2024-08-05 11:28:49 发布

人工智能之浪潮

最新推荐文章于 2024-08-05 11:28:49 发布

阅读量9.7k

点赞数 1

分类专栏： Python实战

本文链接：https://blog.csdn.net/mago2015/article/details/90042323

版权

Python实战专栏收录该内容

24 篇文章 16 订阅

订阅专栏

Python中preprocessing.StandardScaler()在线标准化方法案例

   import numpy as np
    from sklearn import preprocessing
    import utils
    data = utils.readData()
    data0 = np.asarray(data.iloc[:, 2]).reshape(-1, 1)
    data1 = np.asarray(data.iloc[:, 2]).reshape(-1, 1)
    data2 = np.asarray(data.iloc[:, 2]).reshape(-1, 1)

    scaler1 = preprocessing.StandardScaler().fit(data0)
    print("mean  ,  var  : ", (scaler1.mean_, scaler1.scale_, scaler1.var_))

    scaler2 = preprocessing.StandardScaler()
    data2_scaler2 = scaler2.partial_fit(data1)
    print("mean  ,  var  : ", (data2_scaler2.mean_, data2_scaler2.var_))

    scaler3 = preprocessing.StandardScaler()
    data3_scaler3 = scaler3.partial_fit(data2[0:50, :])
    print("mean  ,  var  : ", (data3_scaler3.mean_, data3_scaler3.var_))

    data4_scaler3 = scaler3.partial_fit(data2[50:, :])
    print("mean  ,  var  : ", (data4_scaler3.mean_, data4_scaler3.var_))

    scaler_new = preprocessing.StandardScaler()
    scaler_new.mean_ = data3_scaler3.mean_
    scaler_new.var_ = data3_scaler3.var_
    print("构造新的 scaler 的 mean  ,  var  : ", (scaler_new.mean_, scaler_new.var_))