__________________________ test_rank_deficient_design __________________________
[gw0] linux -- Python 3.7.10 /opt/conda/envs/testvenv/bin/python

    @filterwarnings_normalize
    def test_rank_deficient_design():
        # consistency test that checks that LARS Lasso is handling rank
        # deficient input data (with n_features < rank) in the same way
        # as coordinate descent Lasso
        y = [5, 0, 5]
        for X in ([[5, 0], [0, 5], [10, 10]], [[10, 10, 0], [1e-32, 0, 0], [0, 0, 1]]):
            # To be able to use the coefs to compute the objective function,
            # we need to turn off normalization
            lars = linear_model.LassoLars(0.1, normalize=False)
            coef_lars_ = lars.fit(X, y).coef_
            obj_lars = 1.0 / (2.0 * 3.0) * linalg.norm(
                y - np.dot(X, coef_lars_)
            ) ** 2 + 0.1 * linalg.norm(coef_lars_, 1)
            coord_descent = linear_model.Lasso(0.1, tol=1e-6)
            coef_cd_ = coord_descent.fit(X, y).coef_
            obj_cd = (1.0 / (2.0 * 3.0)) * linalg.norm(
                y - np.dot(X, coef_cd_)
            ) ** 2 + 0.1 * linalg.norm(coef_cd_, 1)
>           assert obj_lars < obj_cd * (1.0 + 1e-8)
E           assert 5.233333333333334 < (0.5711500859559147 * (1.0 + 1e-08))

X          = [[10, 10, 0], [1e-32, 0, 0], [0, 0, 1]]
coef_cd_   = array([0.46399995, 0.        , 4.36999974])
coef_lars_ = array([0.23139881, 0.01860119, 0.        ])
coord_descent = Lasso(alpha=0.1, tol=1e-06)
lars       = LassoLars(alpha=0.1, normalize=False)
obj_cd     = 0.5711500859559147
obj_lars   = 5.233333333333334
y          = [5, 0, 5]

/io/sklearn/linear_model/tests/test_least_angle.py:275: AssertionError