Test failure with sklearn 1.6.0

[penguinpee]

System information
OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Fedora 42 (rawhide)
Sklearn-genetic-opt version: 0.11.1
Scikit-learn version: 1.6.0
Python version: 3.13.1

Describe the bug
Two tests fail with sklearn 1.6.0:

=================================== FAILURES ===================================
____________________________ test_negative_criteria ____________________________
sklearn_genetic/tests/test_feature_selection.py:208: in test_negative_criteria
    evolved_estimator.fit(X_train_b, y_train_b)
sklearn_genetic/genetic_search.py:1138: in fit
    self.scorer_ = check_scoring(self.estimator, self.scoring)
/usr/lib64/python3.13/site-packages/sklearn/utils/_param_validation.py:206: in wrapper
    validate_parameter_constraints(
/usr/lib64/python3.13/site-packages/sklearn/utils/_param_validation.py:98: in validate_parameter_constraints
    raise InvalidParameterError(
E   sklearn.utils._param_validation.InvalidParameterError: The 'scoring' parameter of check_scoring must be a str among {'adjusted_mutual_info_score', 'precision_weighted', 'jaccard_micro', 'f1_samples', 'explained_variance', 'f1_macro', 'neg_median_absolute_error', 'neg_root_mean_squared_error', 'roc_auc_ovo', 'f1_micro', 'rand_score', 'recall', 'jaccard_samples', 'precision', 'neg_mean_absolute_percentage_error', 'neg_negative_likelihood_ratio', 'neg_mean_gamma_deviance', 'recall_samples', 'r2', 'precision_micro', 'neg_mean_absolute_error', 'neg_log_loss', 'neg_mean_squared_error', 'positive_likelihood_ratio', 'recall_weighted', 'f1_weighted', 'normalized_mutual_info_score', 'roc_auc_ovr_weighted', 'precision_macro', 'jaccard_weighted', 'd2_absolute_error_score', 'neg_max_error', 'recall_macro', 'roc_auc_ovr', 'f1', 'top_k_accuracy', 'neg_brier_score', 'roc_auc', 'matthews_corrcoef', 'jaccard_macro', 'accuracy', 'average_precision', 'neg_mean_poisson_deviance', 'neg_mean_squared_log_error', 'neg_root_mean_squared_log_error', 'homogeneity_score', 'fowlkes_mallows_score', 'precision_samples', 'roc_auc_ovo_weighted', 'mutual_info_score', 'completeness_score', 'adjusted_rand_score', 'balanced_accuracy', 'v_measure_score', 'jaccard', 'recall_micro'}, a callable, an instance of 'list', an instance of 'set', an instance of 'tuple', an instance of 'dict' or None. Got 'max_error' instead.
____________________________ test_negative_criteria ____________________________
sklearn_genetic/tests/test_genetic_search.py:277: in test_negative_criteria
    evolved_estimator.fit(X_train_b, y_train_b)
sklearn_genetic/genetic_search.py:531: in fit
    self.scorer_ = check_scoring(self.estimator, self.scoring)
/usr/lib64/python3.13/site-packages/sklearn/utils/_param_validation.py:206: in wrapper
    validate_parameter_constraints(
/usr/lib64/python3.13/site-packages/sklearn/utils/_param_validation.py:98: in validate_parameter_constraints
    raise InvalidParameterError(
E   sklearn.utils._param_validation.InvalidParameterError: The 'scoring' parameter of check_scoring must be a str among {'adjusted_mutual_info_score', 'precision_weighted', 'jaccard_micro', 'f1_samples', 'explained_variance', 'f1_macro', 'neg_median_absolute_error', 'neg_root_mean_squared_error', 'roc_auc_ovo', 'f1_micro', 'rand_score', 'recall', 'jaccard_samples', 'precision', 'neg_mean_absolute_percentage_error', 'neg_negative_likelihood_ratio', 'neg_mean_gamma_deviance', 'recall_samples', 'r2', 'precision_micro', 'neg_mean_absolute_error', 'neg_log_loss', 'neg_mean_squared_error', 'positive_likelihood_ratio', 'recall_weighted', 'f1_weighted', 'normalized_mutual_info_score', 'roc_auc_ovr_weighted', 'precision_macro', 'jaccard_weighted', 'd2_absolute_error_score', 'neg_max_error', 'recall_macro', 'roc_auc_ovr', 'f1', 'top_k_accuracy', 'neg_brier_score', 'roc_auc', 'matthews_corrcoef', 'jaccard_macro', 'accuracy', 'average_precision', 'neg_mean_poisson_deviance', 'neg_mean_squared_log_error', 'neg_root_mean_squared_log_error', 'homogeneity_score', 'fowlkes_mallows_score', 'precision_samples', 'roc_auc_ovo_weighted', 'mutual_info_score', 'completeness_score', 'adjusted_rand_score', 'balanced_accuracy', 'v_measure_score', 'jaccard', 'recall_micro'}, a callable, an instance of 'list', an instance of 'set', an instance of 'tuple', an instance of 'dict' or None. Got 'max_error' instead.
=========================== short test summary info ============================
FAILED sklearn_genetic/tests/test_feature_selection.py::test_negative_criteria - sklearn.utils._param_validation.InvalidParameterError: The 'scoring' parameter of check_scoring must be a str among {'adjusted_mutual_info_score', 'precision_weighted', 'jaccard_micro', 'f1_samples', 'explained_variance', 'f1_macro', 'neg_median_absolute_error', 'neg_root_mean_squared_error', 'roc_auc_ovo', 'f1_micro', 'rand_score', 'recall', 'jaccard_samples', 'precision', 'neg_mean_absolute_percentage_error', 'neg_negative_likelihood_ratio', 'neg_mean_gamma_deviance', 'recall_samples', 'r2', 'precision_micro', 'neg_mean_absolute_error', 'neg_log_loss', 'neg_mean_squared_error', 'positive_likelihood_ratio', 'recall_weighted', 'f1_weighted', 'normalized_mutual_info_score', 'roc_auc_ovr_weighted', 'precision_macro', 'jaccard_weighted', 'd2_absolute_error_score', 'neg_max_error', 'recall_macro', 'roc_auc_ovr', 'f1', 'top_k_accuracy', 'neg_brier_score', 'roc_auc', 'matthews_corrcoef', 'jaccard_macro', 'accuracy', 'average_precision', 'neg_mean_poisson_deviance', 'neg_mean_squared_log_error', 'neg_root_mean_squared_log_error', 'homogeneity_score', 'fowlkes_mallows_score', 'precision_samples', 'roc_auc_ovo_weighted', 'mutual_info_score', 'completeness_score', 'adjusted_rand_score', 'balanced_accuracy', 'v_measure_score', 'jaccard', 'recall_micro'}, a callable, an instance of 'list', an instance of 'set', an instance of 'tuple', an instance of 'dict' or None. Got 'max_error' instead.
FAILED sklearn_genetic/tests/test_genetic_search.py::test_negative_criteria - sklearn.utils._param_validation.InvalidParameterError: The 'scoring' parameter of check_scoring must be a str among {'adjusted_mutual_info_score', 'precision_weighted', 'jaccard_micro', 'f1_samples', 'explained_variance', 'f1_macro', 'neg_median_absolute_error', 'neg_root_mean_squared_error', 'roc_auc_ovo', 'f1_micro', 'rand_score', 'recall', 'jaccard_samples', 'precision', 'neg_mean_absolute_percentage_error', 'neg_negative_likelihood_ratio', 'neg_mean_gamma_deviance', 'recall_samples', 'r2', 'precision_micro', 'neg_mean_absolute_error', 'neg_log_loss', 'neg_mean_squared_error', 'positive_likelihood_ratio', 'recall_weighted', 'f1_weighted', 'normalized_mutual_info_score', 'roc_auc_ovr_weighted', 'precision_macro', 'jaccard_weighted', 'd2_absolute_error_score', 'neg_max_error', 'recall_macro', 'roc_auc_ovr', 'f1', 'top_k_accuracy', 'neg_brier_score', 'roc_auc', 'matthews_corrcoef', 'jaccard_macro', 'accuracy', 'average_precision', 'neg_mean_poisson_deviance', 'neg_mean_squared_log_error', 'neg_root_mean_squared_log_error', 'homogeneity_score', 'fowlkes_mallows_score', 'precision_samples', 'roc_auc_ovo_weighted', 'mutual_info_score', 'completeness_score', 'adjusted_rand_score', 'balanced_accuracy', 'v_measure_score', 'jaccard', 'recall_micro'}, a callable, an instance of 'list', an instance of 'set', an instance of 'tuple', an instance of 'dict' or None. Got 'max_error' instead.
============ 2 failed, 110 passed, 6 deselected in 88.46s (0:01:28) ============

Fedora is in the process of updating NumPy. To make scikit-learn work with numpy 2.2.0 it was updated to version 1.6.0. With that update I'm seeing above test failures.