This function seems to fail often.
def test_list_datasets_with_high_size_parameter(self):
datasets_a = openml.datasets.list_datasets()
datasets_b = openml.datasets.list_datasets(size=np.inf)
> self.assertEqual(len(datasets_a), len(datasets_b))
E AssertionError: 1505 != 1506
My assumption: due to multi threaded testing, in between the two list dataset calls some new datasets have been uploaded.
We should redesign this test case.