小麦产量和水分利用效率预测代码.docxVIP

PAGE2

小麦产量和水分利用效率预测代码

importpandasaspd

fromsklearn.model_selectionimporttrain_test_split,KFold,GridSearchCV

fromsklearn.metricsimportr2_score,mean_squared_error

importmath

importnumpyasnp

importmatplotlib.pyplotasplt

fromsklearn.linear_modelimportRidge#使用Ridge回归模型

#加载Excel文件并提取Sheet1

file_path=C:\\xx\\xx#替换为您的文件路径

df=pd.read_excel(file_path,sheet_name=Sheet1)

#提取特征和目标值

X=df[[Varietiesinformation,irrigationtreatments,Rinitial,R90,R50,RTm]]#特征列

y=df[GY,WUE]#目标列

#分割数据为训练集和测试集

X_train,X_test,y_train,y_test=train_test_split(

X,y,test_size=0.2,random_state=42)

#定义Ridge回归模型

ridge_model=Ridge()

#定义参数网格，用于搜索最佳alpha值

param_grid={alpha:[0.001,0.01,0.1,1,10,100,1000]}

#使用GridSearchCV进行参数搜索，结合五折交叉验证

grid_search=GridSearchCV(estimator=ridge_model,param_grid=param_grid,cv=5,scoring=neg_mean_squared_error)

grid_search.fit(X_train,y_train)

#获取最佳模型和最佳参数

best_ridge_model=grid_search.best_estimator_

best_alpha=grid_search.best_params_[alpha]

print(f最佳alpha值:{best_alpha})

#使用最佳模型进行预测

y_train_pred=best_ridge_model.predict(X_train)

y_test_pred=best_ridge_model.predict(X_test)

#五折交叉验证

kf=KFold(n_splits=5,shuffle=True,random_state=42)

r2_cv=[]

rmse_cv=[]

rmse_percent_cv=[]

r2_train_cv=[]

rmse_train_cv=[]

rmse_train_percent_cv=[]

#用于保存所有样本的预测值

all_y_pred=np.zeros(len(y))#初始化一个数组来保存所有预测值

#初始化列表以保存真实值和预测值

y_test_list=[]

y_pred_list=[]

forfold,(train_index,test_index)inenumerate(kf.split(X),1):

X_train_fold,X_test_fold=X.iloc[train_index],X.iloc[test_index]

y_train_fold,y_test_fold=y.iloc[train_index],y.iloc[test_index]

#构建Ridge回归模型，使用最佳alpha值

ridge_model_fold=Ridge(alpha=best_alpha)

ridge_model_fold.fit(X_train_fold,y_train_fold)

#预测训练集和测试集

y_train_pred_fold=ridge_model_fold.predict(X_train_fold)#训练集预测

y_test_pred_fold=ridge_model_fold.predict(X_test_fold)#测试集预测

#将真实值和预测值添加到列表

y_test_list.extend(y_test_fold)