报错信息

C:\Python27\python.exe D:/code/ML5/part08/xin/BugTest.py
Traceback (most recent call last):
  File "D:/code/ML5/part08/xin/BugTest.py", line 28, in <module>
    class_names=["is_date:no","is_date:yes"],filled=True,rounded=True)
  File "C:\Python27\lib\site-packages\sklearn\tree\export.py", line 392, in export_graphviz
    check_is_fitted(decision_tree, 'tree_')
  File "C:\Python27\lib\site-packages\sklearn\utils\validation.py", line 762, in check_is_fitted
    raise TypeError("%s is not an estimator instance." % (estimator))
TypeError: <sklearn.tree._tree.Tree object at 0x0000000018479D98> is not an estimator instance.

出问题的代码

dtc.fit(X_train,y_train)

#6.可视化
from sklearn.tree import export_graphviz
export_graphviz(dtc.tree_,out_file="tan1.dot",)

上文中的export_graphviz方法，参数是 dtc.tree_ ，这里取不到这个值。 需要给dtc.fit(X_train,y_train)赋一个返回值。

改成这样

dtc_tree=dtc.fit(X_train,y_train)

#6.可视化
from sklearn.tree import export_graphviz
export_graphviz(dtc_tree,out_file="tan1.dot",)