[ML] Scikit Learn : Pipeline

Pipeline

- pipeline 이란?

• test_train_split() 은 아니다

- 코드로 구현하기

from sklearn.pipeline import Pipeline
from sklearn.tree import DecisionTreeClassifier
from sklearn.preprocessing import StandardScaler

estimators = [
    ('scaler', StandardScaler()),
    ('clf', DecisionTreeClassifier())
]

pipe = Pipeline(estimators)

- 구조확인하기

• steps

pipe.steps

>>>
[('scaler', StandardScaler()), ('clf', DecisionTreeClassifier())]

• 스텝별 객체 호출

pipe.steps[0]
>>>
('scaler', StandardScaler())

- 파라미터 지정

• set_params

pipe.set_params(clf__max_depth = 2)
pipe.set_params(clf__random_state =13)

• 스탭이름 + 언더바 두개 (__) + 파라미터 값

- 학습 및 평가

from sklearn.model_selection import train_test_split

X_train, X_test, y_train ,y_test = train_test_split(X,y, test_size=0.2, random_state=13,
                                                    stratify = y)

pipe.fit(X_train, y_train)

from sklearn.metrics import accuracy_score

y_pred_train = pipe.predict(X_train)
y_pred_test = pipe.predict(X_test)

print(f'Train acc : {accuracy_score(y_pred_train , y_train)}')
print(f'Test acc : {accuracy_score(y_pred_test , y_test)}')