机器学习---数据预处理

import numpy as np
from sklearn import preprocessing

data = np.array([[1,2,34,4],[3,43,53,4],[23,2,2,466]])
data_standard = preprocessing.scale(data)
print("标准化：n",data_standard)
print("特征均值：n",data_standard.mean(axis=0))
print("特征标准差：n",data_standard.std(axis=0))

data_scaler = preprocessing.MinMaxScaler(feature_range=(0,1))
data_sacled = data_scaler.fit_transform(data)
print("缩放后：n",data_sacled)


data_nomalized = preprocessing.normalize(data,norm="l2",axis=0)
print("L1范数归一化：n",data_nomalized)


data_binarized = preprocessing.Binarizer(threshold=3).transform(data)
print("二值化：n",data_binarized)

data_encoder = preprocessing.OneHotEncoder()
data_encoder.fit(data)
data_encoded = data_encoder.transform(X=[[1,2,34,466]]).toarray()
print("ont-hot编码：n",data_encoded)

label = ["足球","篮球","足球","排球","网球","篮球"]
laber_encoder= preprocessing.LabelEncoder()
laber_encoder.fit(label)
for i,item in enumerate(laber_encoder.classes_,1):
    print(item,"----->",i)
label_ = ["足球","排球","网球"]
label_encoded = laber_encoder.transform(label_)
print("encodered_label=",label_encoded)
label_decoded = laber_encoder.inverse_transform([3,1,0,2])
print(label_decoded)