youchounobb/data-mining-introduction

数据挖掘入门介绍

data-mining-introduction's Issues

预处理部分 df=sparse.hstack((item_category_list,df)) 跑不了

报类型错误

Undefined names

flake8 testing of https://github.com/YouChouNoBB/sklearn-introduction on Python 3.6.3

$ flake8 . --count --select=E901,E999,F821,F822,F823 --show-source --statistics

./blending.py:28:36: F821 undefined name 'preprocess_train_input'
    train_x_id, train_x, train_y = preprocess_train_input()
                                   ^
./blending.py:29:30: F821 undefined name 'preprocess_val_input'
    val_x_id, val_x, val_y = preprocess_val_input()
                             ^
./ensemble.py:15:36: F821 undefined name 'train_test_split'
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=111)
                                   ^
./ensemble.py:21:49: F821 undefined name 'RandomForestRegressor'
model_rf = Regressor(dataset=dataset, estimator=RandomForestRegressor, parameters={'n_estimators': 50},name='rf')
                                                ^
./ensemble.py:22:49: F821 undefined name 'LinearRegression'
model_lr = Regressor(dataset=dataset, estimator=LinearRegression, parameters={'normalize': True},name='lr')
                                                ^
./ensemble.py:30:49: F821 undefined name 'LinearRegression'
stacker = Regressor(dataset=stack_ds, estimator=LinearRegression)
                                                ^
./ensemble.py:33:40: F821 undefined name 'mean_absolute_error'
results = stacker.validate(k=10,scorer=mean_absolute_error)
                                       ^
./ensemble.py:42:36: F821 undefined name 'train_test_split'
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=111)
                                   ^
./ensemble.py:48:49: F821 undefined name 'RandomForestRegressor'
model_rf = Regressor(dataset=dataset, estimator=RandomForestRegressor, parameters={'n_estimators': 50},name='rf')
                                                ^
./ensemble.py:49:49: F821 undefined name 'LinearRegression'
model_lr = Regressor(dataset=dataset, estimator=LinearRegression, parameters={'normalize': True},name='lr')
                                                ^
./ensemble.py:57:49: F821 undefined name 'LinearRegression'
stacker = Regressor(dataset=stack_ds, estimator=LinearRegression)
                                                ^
./ensemble.py:60:40: F821 undefined name 'mean_absolute_error'
results = stacker.validate(k=10,scorer=mean_absolute_error)
                                       ^
./ensemble.py:65:32: F821 undefined name 'boston_dataset'
dataset = Dataset(preprocessor=boston_dataset)
                               ^
./ensemble.py:67:49: F821 undefined name 'RandomForestRegressor'
model_rf = Regressor(dataset=dataset, estimator=RandomForestRegressor, parameters={'n_estimators': 151},name='rf')
                                                ^
./ensemble.py:68:49: F821 undefined name 'LinearRegression'
model_lr = Regressor(dataset=dataset, estimator=LinearRegression, parameters={'normalize': True},name='lr')
                                                ^
./ensemble.py:69:50: F821 undefined name 'KNeighborsRegressor'
model_knn = Regressor(dataset=dataset, estimator=KNeighborsRegressor, parameters={'n_neighbors': 15},name='knn')
                                                 ^
./ensemble.py:73:33: F821 undefined name 'mean_absolute_error'
weights = pipeline.find_weights(mean_absolute_error)
                                ^
17    F821 undefined name 'preprocess_train_input'
17

heamy中训练集的x如果传进去稀疏矩阵会报错，请问该怎么解决

if self._X_train.shape[0] != self._y_train.shape[0]:

会报这样的错，谢谢！

对于blending和stacking的理解

您好，我最近看了您在天池的视频：sklearn入门系列三：数据预处理、模型评估、模型融合。其中在讲到blending&stacking时，您参照的是模型融合这篇博客的内容。但是我在KAGGLE ENSEMBLING GUIDE中，发现blending&stacking的解释（具体位置在Stacked generalization和Blending下）和那篇博客的内容正好相反。我也查了一下网上的内容，对blending&stacking的解释也是分为这两种。因此想请教您一下，blending&stacking这两个概念应该参照哪种解释？