fromsklearn.linear_modelimportElasticNet,Lasso,BayesianRidge,LassoLarsICfromsklearn.ensembleimportRandomForestRegressor,GradientBoostingRegressorfromsklearn.kernel_ridgeimportKernelRidgefromsklearn.pipelineimportmake_pipelinefromsklearn.preprocessingimportRobustScalerfromsklearn.baseimportBaseEstimator,TransformerMixin,RegressorMixin,clonefromsklearn.model_selectionimportKFold,cross_val_score,train_test_splitfromsklearn.metricsimportmean_squared_errorimportxgboostasxgbimportlightgbmaslgb

#Validation function
n_folds=5defrmsle_cv(model):kf=KFold(n_folds,shuffle=True,random_state=42).get_n_splits(train.values)#シャッフル
rmse=np.sqrt(-cross_val_score(model,train.values,y_train,scoring="neg_mean_squared_error",cv=kf))return(rmse)defrmsle(y,y_pred):returnnp.sqrt(mean_squared_error(y,y_pred))

classStackingAveragedModels(BaseEstimator,RegressorMixin,TransformerMixin):def__init__(self,base_models,meta_model,n_folds=5):self.base_models=base_modelsself.meta_model=meta_modelself.n_folds=n_folds# モデルのクローンにデータをfitさせる
deffit(self,X,y):self.base_models_=[list()forxinself.base_models]self.meta_model_=clone(self.meta_model)kfold=KFold(n_splits=self.n_folds,shuffle=True,random_state=156)# 学習ベースモデルがフォールド外予測をします
# そのフォールド外予測はメタモデルで必要となります
out_of_fold_predictions=np.zeros((X.shape[0],len(self.base_models)))fori,modelinenumerate(self.base_models):fortrain_index,holdout_indexinkfold.split(X,y):instance=clone(model)self.base_models_[i].append(instance)instance.fit(X[train_index],y[train_index])y_pred=instance.predict(X[holdout_index])out_of_fold_predictions[holdout_index,i]=y_pred# フォールド外予測を用いてメタモデルの学習
self.meta_model_.fit(out_of_fold_predictions,y)returnself#全てのベースモデルの予測値と予測の平均値をメタ特徴として、メタモデルで最終予測を行います!
defpredict(self,X):meta_features=np.column_stack([np.column_stack([model.predict(X)formodelinbase_models]).mean(axis=1)forbase_modelsinself.base_models_])returnself.meta_model_.predict(meta_features)

stacked_averaged_models=StackingAveragedModels(base_models=(ENet,GBoost,KRR),meta_model=lasso)score=rmsle_cv(stacked_averaged_models)print("Stacking Averaged models score: {:.4f} ({:.4f})".format(score.mean(),score.std()))

Stacking Averaged models score: 0.1085 (0.0074)

stacked_averaged_models.fit(train.values,y_train)stacked_train_pred=stacked_averaged_models.predict(train.values)stacked_pred=np.expm1(stacked_averaged_models.predict(test.values))print(rmsle(y_train,stacked_train_pred))

0.0781571937916

#学習
model_xgb.fit(train,y_train)xgb_train_pred=model_xgb.predict(train)#テスト
xgb_pred=np.expm1(model_xgb.predict(test))#評価
print(rmsle(y_train,xgb_train_pred))

0.0785165142425

model_lgb.fit(train,y_train)lgb_train_pred=model_lgb.predict(train)lgb_pred=np.expm1(model_lgb.predict(test.values))print(rmsle(y_train,lgb_train_pred))

0.0716757468834

'''RMSE on the entire Train data when averaging'''print('RMSLE score on train data:')print(rmsle(y_train,stacked_train_pred*0.70+xgb_train_pred*0.15+lgb_train_pred*0.15))

RMSLE score on train data:
0.0752190464543

ensemble=stacked_pred*0.70+xgb_pred*0.15+lgb_pred*0.15

sub=pd.DataFrame()sub['Id']=test_IDsub['SalePrice']=ensemblesub.to_csv('submission.csv',index=False)