1
$\begingroup$

I find this really weird, and the code is really straight forward. What am I doing wrong ?

from sklearn.model_selection import GridSearchCV
scoring_type="accuracy"
preprocess_data(X,y,0)
p_grid = {'min_samples_split':np.arange(2,10),'min_samples_leaf': np.arange(1,10)}
Tree_opt = GridSearchCV(estimator=DecisionTreeClassifier(random_state=42), param_grid=p_grid, scoring=scoring_type, cv=10)
Tree_opt.fit(X_train,y_train)

print("Best training params: {}".format(Tree_opt.best_params_))
print("Best training Score: {}".format(Tree_opt.best_score_))
dot_data = tree.export_graphviz(Tree_opt, out_file=None,feature_names=labels,class_names=class_names,filled=True, rounded=True,special_characters=True)
graph = graphviz.Source(dot_data)

graph
print(Tree_opt.score(X_test, y_test))

And I'm getting the following error:

NotFittedError                            Traceback (most recent call last)
<ipython-input-14-1c7fa906f99b> in <module>
      8 print("Best training params: {}".format(Tree_opt.best_params_))
      9 print("Best training Score: {}".format(Tree_opt.best_score_))
---> 10 dot_data = tree.export_graphviz(Tree_opt, out_file=None,feature_names=labels,class_names=class_names,filled=True, rounded=True,special_characters=True)
     11 graph = graphviz.Source(dot_data)
     12 

~\Anaconda3\lib\site-packages\sklearn\tree\export.py in export_graphviz(decision_tree, out_file, max_depth, feature_names, class_names, label, filled, leaves_parallel, impurity, node_ids, proportion, rotate, rounded, special_characters, precision)
    394                 out_file.write('%d -> %d ;\n' % (parent, node_id))
    395 
--> 396     check_is_fitted(decision_tree, 'tree_')
    397     own_file = False
    398     return_string = False

~\Anaconda3\lib\site-packages\sklearn\utils\validation.py in check_is_fitted(estimator, attributes, msg, all_or_any)
    949 
    950     if not all_or_any([hasattr(estimator, attr) for attr in attributes]):
--> 951         raise NotFittedError(msg % {'name': type(estimator).__name__})
    952 
    953 

NotFittedError: This GridSearchCV instance is not fitted yet. Call 'fit' with appropriate arguments before using this method.

I also have this warning I don't think it's important:

C:\Users\Flow\Anaconda3\lib\site-packages\sklearn\model_selection\_search.py:841: DeprecationWarning: The default of the `iid` parameter will change from True to False in version 0.22 and will be removed in 0.24. This will change numeric results when test-set sizes are unequal.
  DeprecationWarning)

I think the fit works because I can get the first prints as follows:

Best training params: {'min_samples_leaf': 3, 'min_samples_split': 2}
Best training Score: 0.8809523809523809
$\endgroup$
0
$\begingroup$

Try using Tree_opt.best_estimator_. The grid search object passes fit and score functions through to the best estimator when refit=True, but other (esp. more specific) methods need to be called directly from best estimator object.

$\endgroup$
0
0
$\begingroup$

Here is the modified correct code :

from sklearn.model_selection import GridSearchCV
scoring_type="accuracy"
preprocess_data(X,y,0)
p_grid = {'min_samples_split':np.arange(2,10),'min_samples_leaf': np.arange(1,10)}
Tree_opt = GridSearchCV(estimator=DecisionTreeClassifier(random_state=42), param_grid=p_grid, scoring=scoring_type, cv=10, refit=True)
Tree_opt.fit(X_train,y_train)


print("Best training params: {}".format(Tree_opt.best_params_))
print("Best training Score: {}".format(Tree_opt.best_score_))
print("Best training Score: {}".format(Tree_opt.best_estimator_))
Tree_opt.best_estimator_.fit(X_train, y_train)
dot_data = tree.export_graphviz(Tree_opt.best_estimator_, out_file=None,feature_names=labels,class_names=class_names,filled=True, rounded=True,special_characters=True)
graph = graphviz.Source(dot_data)

graph

Indeed we need to call on .best_estimator_ instance

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.