Python，功能选择

Question

When I use RFE to select the most important features in my data set, it returns all features , instead of returning the number of features that I specified

here is simple code:

  from sklearn.svm import SVC
  from sklearn.datasets import load_digits
  from sklearn.feature_selection import RFE
  import matplotlib.pyplot as plt
  new_X=np.array([[1,2,3,3],[1,2,4,4],[3,1,3,4],[3,1,4,5]])
  new_Y=np.array([1,1,0,0])
  svc = SVC(kernel="linear", C=1)
  rfe = RFE(estimator=svc, n_features_to_select=2, step=1)
  rfe.fit(new_X, new_Y)
  ranking = rfe.ranking_
  len(ranking)

Answer 1

You are looking at the wrong thing.

rfe.ranking_ will always return the ranking of all the features. But the selected features have value 1 in ranking_

Print the ranking_ and see:

ranking
# Output: array([1, 1, 3, 2])

This means that 1st and 2nd features are selected.