Python - 最小偏二乘法PLS代码

# -*-coding:GBK -*-
import sys
import math
import xlrd 
import xlwt  
import xlutils 
import data_pre
import xlsxwriter
import numpy as np
from sklearn import svm
from sklearn import preprocessing
from sklearn import model_selection
from sklearn.impute import SimpleImputer
from sklearn.metrics import accuracy_score
from sklearn.model_selection import GridSearchCV
from sklearn.model_selection import train_test_split
from sklearn.cross_decomposition import PLSRegression

def PLS():
    for tag in [10,20,30,50,100,200,300]:
        x, y = data_pre.load('data_test_mean_mean.csv')
        x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=1, train_size=0.7)
        pls2 = PLSRegression(n_components = tag)
        pls2.fit(x_train, y_train.ravel())
        y_test_pre = pls2.predict(x_test)
        sumd,mxd,cnt = 0.0,0.0,0.0
        for i in range (len(y_test_pre)):
            if(math.fabs(y_test_pre[i]-y_test[i])>1):
                cnt+=1
            sumd+=math.fabs(y_test_pre[i]-y_test[i])
            mxd = max(mxd,math.fabs(y_test_pre[i]-y_test[i]))
        sumd/=len(y_test_pre)
        cnt/=len(y_test_pre)
        accuracy =pls2.score(x_test, y_test)
        print('{:.2f},{:.2f},{:.2f},{:.2f}%'.format(accuracy,sumd,mxd,cnt*100))

if __name__ == '__main__':
    PLS()