Sklearn警告解决办法: UserWarning: X has feature names, but MLPClassifier was fitted without feature names

警告出现于在使用sklearn中的MLPClassifier（多层神经网络分类器）中
完整警告信息：

/usr/local/lib/python3.9/site-packages/sklearn/base.py:443: UserWarning: X has feature names, but MLPClassifier was fitted without feature names
warnings.warn(

警告部分代码：

from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import StandardScaler

# Scale the data
def scale(x):
    scaler = StandardScaler()  
    scaler.fit(x) 
    x = scaler.transform(x)
    return x

# 3 Multi-layer Perceptron Classifier
def MLPClf(x, y):
    x = scale(x)
    clf = MLPClassifier(solver='sgd', alpha=1e-5, max_iter=400, hidden_layer_sizes=(5,), random_state=1)
    clf = clf.fit(x, y)
    return clf

分析和解决办法：因为MLPClassifier需要包含feature names的输入变量，但是x由于经过了StandardScaler()的feature scaling，导致其被转化为了array格式，也就不存在feature names了。因此解决方法也很直观，再把scaling后的x转回dataframe就可以了，唯一要注意的是要提前把columns储存起来。
修改后代码：

from sklearn.neural_network import MLPClassifier
from sklearn.preprocessing import StandardScaler

# Scale the data
def scale(x):  
    # store columns in advance
    cols = x.columns
    scaler = StandardScaler()  
    scaler.fit(x) 
    x = scaler.transform(x)
    # avoid warning, transform it back to df
    x = pd.DataFrame(x, columns=cols)
    return x

# 3 Multi-layer Perceptron Classifier
def MLPClf(x, y):
    x = scale(x)
    clf = MLPClassifier(solver='sgd', alpha=1e-5, max_iter=400, hidden_layer_sizes=(5,), random_state=1)
    clf = clf.fit(x, y)
    return clf