RandomForestClassifier(随机森林检测每个特征的重要性及每个样例属于哪个类的概率)

#In the next recipe, we'll look at how to tune the random forest classifier.
#Let's start by importing datasets:

from sklearn import datasets
X, y = datasets.make_classification(1000)

# X(1000,20)
#y(1000) 取值范围【0,1】

from sklearn.ensemble import RandomForestClassifier
rf = RandomForestClassifier()
rf.n_jobs=-1

rf.fit(X, y)
print ("Accuracy:	", (y == rf.predict(X)).mean())
print ("Total Correct:	", (y == rf.predict(X)).sum())


#每个例子属于哪个类的概率
probs = rf.predict_proba(X)
import pandas as pd
probs_df = pd.DataFrame(probs, columns=['0', '1'])
probs_df['was_correct'] = rf.predict(X) == y
import matplotlib.pyplot as plt
f, ax = plt.subplots(figsize=(7, 5))
probs_df.groupby('0').was_correct.mean().plot(kind='bar', ax=ax)
ax.set_title("Accuracy at 0 class probability")
ax.set_ylabel("% Correct")
ax.set_xlabel("% trees for 0")
f.show()

#检测重要特征
rf = RandomForestClassifier()
rf.fit(X, y)
f, ax = plt.subplots(figsize=(7, 5))
ax.bar(range(len(rf.feature_importances_)),rf.feature_importances_)
ax.set_title("Feature Importances")
f.show()