import numpy as np
from sklearn.base import BaseEstimator, RegressorMixin
from sklearn.metrics._scorer import _BaseScorer, _MultimetricScorer

# Step 1: Define a simple estimator
class SimpleEstimator(BaseEstimator, RegressorMixin):
    def fit(self, X, y):
        self.mean_ = np.mean(y)
        return self

    def predict(self, X):
        return np.full(X.shape[0], self.mean_)

# Step 2: Define a custom scorer inheriting from _BaseScorer and a function which estimates score
class SimpleScorer(_BaseScorer):
    def _score(self, method_caller, estimator, X, y_true, sample_weight=None):
        y_pred = method_caller(estimator, "predict", X)
        return self._score_func(y_true, y_pred, **self._kwargs)

def default_score(estimator, X, y, sample_weight=None, **kws):
    return estimator.score(X, y, sample_weight=sample_weight)

def mse(y, y_pred):
    return np.mean((y - y_pred)**2)

# Step 3: Create a _MultimetricScorer with multiple scorers
scorers = {
    "mse": SimpleScorer(mse, sign=1, kwargs={}),
    "default": default_score
}
multi_scorer = _MultimetricScorer(scorers=scorers)

# Step 4: Generate sample data
X = np.random.rand(10, 2)
y = np.random.rand(10)
sample_weight = np.random.rand(10)

# Step 5: Fit the estimator and evaluate using the multi-scorer
estimator = SimpleEstimator()
estimator.fit(X, y)
scores = multi_scorer(estimator, X, y, sample_weight=sample_weight)

print("Scores:", scores)

AttributeError                            Traceback (most recent call last)
Cell In[6], line 4
      2 estimator = SimpleEstimator()
      3 estimator.fit(X, y)
----> 4 scores = multi_scorer(estimator, X, y, sample_weight=sample_weight)
      6 print("Scores:", scores)

File ~/anaconda3/envs/scoring-311-2/lib/python3.11/site-packages/sklearn/metrics/_scorer.py:144, in _MultimetricScorer.__call__(self, estimator, *args, **kwargs)
    142     if "sample_weight" in kwargs:
    143         for name, scorer in self._scorers.items():
--> 144             if scorer._accept_sample_weight():
    145                 routed_params[name].score["sample_weight"] = kwargs[
    146                     "sample_weight"
    147                 ]
    149 for name, scorer in self._scorers.items():

AttributeError: 'function' object has no attribute '_accept_sample_weight'