Describe the bug

I wanted to learn more about how cross_validation is implemented in sklearn and came across what I think is a faulty or confusing test.

The test test_cross_validate in model_selection/tests/test_validation.py creates two datasets, one regression model and one classification model. The test loops over the dataset x model combinations and fits and scores the evaluation metrics. In the loop est is used to represent the model under consideration (Lasso or SVC). However, in that process, clone(reg).fit(...) is called, so that the Lasso regression is fit and evaluated in both iterations, also for the classification data. If I'm not mistaken, this should be clone(est).fit(...), to train and evaluate the SVC classifier.

However, using clone on est actually fails the test, which I did not expect given the comment "It's okay to evaluate regression metrics on classification too".

To me this shows that the test is not testing what it is supposed to be testing.

Steps/Code to Reproduce

Run the test test_cross_validate()

Expected Results

The test should use the regressor on the regression data and the classier on the classification data, while passing the test.

Actual Results

The test uses the Lasso regressor on both the regression data and the classification data, while passing the test.

Versions

System:
    python: 3.9.15 | packaged by conda-forge | (main, Nov 22 2022, 08:41:22) [MSC v.1929 64 bit (AMD64)]
executable: C:\Users\s.elgewily\Miniconda3\envs\sklearn-env2\python.exe
   machine: Windows-10-10.0.19044-SP0
Python dependencies:
      sklearn: 1.3.dev0
          pip: 22.3.1
   setuptools: 66.0.0
        numpy: 1.24.1
        scipy: 1.10.0
       Cython: 0.29.33
       pandas: None
   matplotlib: None
       joblib: 1.2.0
threadpoolctl: 3.1.0
Built with OpenMP: True
threadpoolctl info:
       user_api: blas
   internal_api: mkl
         prefix: libblas
       filepath: C:\Users\s.elgewily\Miniconda3\envs\sklearn-env2\Library\bin\libblas.dll
        version: 2022.1-Product
threading_layer: intel
    num_threads: 4
       user_api: openmp
   internal_api: openmp
         prefix: vcomp
       filepath: C:\Users\s.elgewily\Miniconda3\envs\sklearn-env2\vcomp140.dll
        version: None
    num_threads: 8