Sklearn 1.6.0 compatibility

[SvenKlaassen]

With scikit-learn 1.6.0, __sklearn_tags__ were introduces, see Estimator Tags
This raised some issues at

• lightgbm: [python-package] simplify scikit-learn 1.6+ tags support microsoft/LightGBM#6735 [python-package][R-package] adapt to scikit-learn 1.6 testing changes, pin more packages in R 3.6 CI jobs microsoft/LightGBM#6718
• xgboost: Incompability between scikit-learn and xgboost dmlc/xgboost#11093

Further, this caused issues for stacked Global Learners as in Example Gallery

Minimal Example

import doubleml as dml
import sklearn
import numpy as np
import pandas as pd
from sklearn.linear_model import LassoCV
from doubleml.rdd.datasets import make_simple_rdd_data
from doubleml.rdd import RDFlex
from doubleml.utils.global_learner import GlobalRegressor
from sklearn.ensemble import StackingRegressor

print(sklearn.__version__)
print(dml.__version__)

np.random.seed(42)
data_dict = make_simple_rdd_data(n_obs=1000, fuzzy=False)
cov_names = ['x' + str(i) for i in range(data_dict['X'].shape[1])]
df = pd.DataFrame(np.column_stack((data_dict['Y'], data_dict['D'], data_dict['score'], data_dict['X'])), columns=['y', 'd', 'score'] + cov_names)
dml_data = dml.DoubleMLData(df, y_col='y', d_cols='d', x_cols=cov_names, s_col='score')
ml_g = StackingRegressor([("global", GlobalRegressor(LassoCV())),
                          ("local", LassoCV())], final_estimator=LassoCV())
rdflex_obj = RDFlex(dml_data, ml_g, fuzzy=False)
rdflex_obj.fit()

Output:

1.6.0
0.9.3
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
[<ipython-input-4-deff49535681>](https://localhost:8080/#) in <cell line: 0>()
     20                           ("local", LassoCV())], final_estimator=LassoCV())
     21 rdflex_obj = RDFlex(dml_data, ml_g, fuzzy=False)
---> 22 rdflex_obj.fit()

6 frames
[/usr/local/lib/python3.11/dist-packages/sklearn/ensemble/_base.py](https://localhost:8080/#) in _validate_estimators(self)
    232         for est in estimators:
    233             if est != "drop" and not is_estimator_type(est):
--> 234                 raise ValueError(
    235                     "The estimator {} should be a {}.".format(
    236                         est.__class__.__name__, is_estimator_type.__name__[3:]

ValueError: The estimator GlobalRegressor should be a regressor.

[SvenKlaassen]

The PR #291 updates global learners to work with scikit-learn>=1.6.0
The issue will be closed if also xgboost is updated.