CODEWITHSUNDEEP
machine-learningscikit-learnadaboost
Joed Ngangmeni
Joed Ngangmeni

Reputation: 13

'AdaBoostClassifier' object has no attribute 'estimator_'

I am trying to time each round of an adaboost algorithm (how long each additional tree takes to build). I conda installed scikit-learn 1.4.0 (because on their website it says this version) and all the other reqs to run code.

Here is my code:

Y, z = parse.getHARData() #returns my features Y and labels z

Z_train, Z_test, j_train, j_test = train_test_split(Y, z, test_size=0.30, shuffle=True)

b_estimator = DecisionTreeClassifier(max_depth=DEPTH)

ada = AdaBoostClassifier(estimator=b_estimator, n_estimators=NUMTREES)

elapsed_times = []

for stage in range(NUMTREES): start_time = time.time()

    # Access and fit the current base estimator
    base_estimator = ada._make_estimator(append=True, random_state=42)
    base_estimator.fit(Z_train, j_train)
    
    elapsed_time = time.time() - start_time
    elapsed_times.append(elapsed_time)

I was expecting this to start the timer, grow a tree using the forest's previous information, add that tree to the ensemble, stop the time, and append the elapsed time to elapsed_times.

instead, its returning this error:

AttributeError                            Traceback (most recent call last)
Cell In[9], line 7
      4 start_time = time.time()
      6 # Access and fit the current base estimator
----> 7 base_estimatr = ada._make_estimator(append=True, random_state=42)
      8 base_estimatr.fit(Z_train, j_train)
     10 elapsed_time = time.time() - start_time

File ~/anaconda3/envs/ADA/lib/python3.11/site-packages/sklearn/ensemble/_base.py:141, in BaseEnsemble._make_estimator(self, append, random_state)
    135 def _make_estimator(self, append=True, random_state=None):
    136     """Make and configure a copy of the `estimator_` attribute.
    137 
    138     Warning: This method should be used to properly instantiate new
    139     sub-estimators.
    140     """
--> 141     estimator = clone(self.estimator_)
    142     estimator.set_params(**{p: getattr(self, p) for p in self.estimator_params})
    144     if random_state is not None:

AttributeError: 'AdaBoostClassifier' object has no attribute 'estimator_'

Upvotes: 0

Views: 227

Answers (1)

Ben Reiniger
Ben Reiniger

Reputation: 12738

The attribute estimator_ gets set early in the fit method:

https://github.com/scikit-learn/scikit-learn/blob/9e38cd00d032f777312e639477f1f52f3ea4b3b7/sklearn/ensemble/_weight_boosting.py#L149

https://github.com/scikit-learn/scikit-learn/blob/9e38cd00d032f777312e639477f1f52f3ea4b3b7/sklearn/ensemble/_base.py#L125

So you can just manually call ada._validate_estimator() before your loop to initialize it, and this particular error should be fixed.

But you're skipping a lot in your loop, most notably the weight updates. You should at least call _boost, and check the source to see if there's anything else to include in the loop. (It's unfortunate there's no partial_fit for adaboost...). Or, consider just modifying your local install of sklearn to include the timing you want to measure.

Upvotes: 1

Related Questions