pipeline(管道的连续应用)

# -*- coding: utf-8 -*-
"""
Created on Tue Aug 09 22:55:06 2016

@author: Administrator
"""
#方法1
from sklearn import svm
from sklearn.datasets import samples_generator
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import f_regression
from sklearn.pipeline import Pipeline

# 生成数据
X, y = samples_generator.make_classification(n_informative=5, n_redundant=0, random_state=42)

# 定义Pipeline，先方差分析，再SVM
anova_filter = SelectKBest(f_regression, k=5)
clf = svm.SVC(kernel='linear')
pipe = Pipeline([('anova', anova_filter), ('svc', clf)])

# 设置anova的参数k=10，svc的参数C=0.1（用双下划线"__"连接！）
pipe.set_params(anova__k=10, svc__C=.1)
pipe.fit(X, y)

prediction = pipe.predict(X) #管道怎么会预测，见文章末尾

pipe.score(X, y)                        

# 得到 anova_filter 选出来的特征
s = pipe.named_steps['anova'].get_support()
print(s)

#方法2
import numpy as np

from sklearn import linear_model, decomposition, datasets
from sklearn.pipeline import Pipeline
from sklearn.grid_search import GridSearchCV

digits = datasets.load_digits()
X_digits = digits.data
y_digits = digits.target

# 定义管道，先降维(pca)，再逻辑回归
pca = decomposition.PCA()
logistic = linear_model.LogisticRegression()
pipe = Pipeline(steps=[('pca', pca), ('logistic', logistic)])

# 把管道再作为grid_search的estimator
n_components = [20, 40, 64]
Cs = np.logspace(-4, 4, 3)
estimator = GridSearchCV(pipe, dict(pca__n_components=n_components, logistic__C=Cs))

estimator.fit(X_digits, y_digits)

#Pipeline 无预测函数,他用管道中最后一个预测函数

Applies transforms to the data, and the predict method of the final estimator. Valid only if the final estimator implements predict.