[BUG] sporadic failure of `test_trafo_1` at timings check

[fkiraly]

test_trafo_1 fails sporadically with a timings check. Example traceback below.

If this is not a strict requirement, we should perhaps remove this or make it more lenient?

================================== FAILURES ===================================
______________ test_trafo_1[search_space2-memory_warm_start2-4] _______________

random_state = 4
search_space = {'min_samples_split': [np.int64(22), np.int64(23), np.int64(24), np.int64(25), np.int64(26), np.int64(27), ...]}
memory_warm_start =    min_samples_split  score
0                 22     22
1                 23     23
2                 24     24
3     ...     27
6                 28     28
7                 29     29
8                 30     30
9                 31     31

    @pytest.mark.parametrize("random_state", random_state_para_0)
    @pytest.mark.parametrize("search_space, memory_warm_start", search_space_para_1)
    def test_trafo_1(random_state, search_space, memory_warm_start):
        search_space = search_space
        memory_warm_start = memory_warm_start
    
        c_time_0 = time.perf_counter()
        hyper = Hyperactive()
        hyper.add_search(
            objective_function_1,
            search_space,
            n_iter=10,
            random_state=random_state,
            initialize={"random": 1},
        )
        hyper.run()
        d_time_0 = time.perf_counter() - c_time_0
    
        c_time_1 = time.perf_counter()
        hyper = Hyperactive()
        hyper.add_search(
            objective_function_1,
            search_space,
            n_iter=10,
            random_state=random_state,
            initialize={"random": 1},
            memory_warm_start=memory_warm_start,
        )
        hyper.run()
        d_time_1 = time.perf_counter() - c_time_1
    
>       assert d_time_1 < d_time_0 * 0.5
E       assert 1.027204499999982 < (1.9211578000000031 * 0.5)

[SimonBlanke]

We should remove this test. It is one of several tests, that work with measuring and comparing optimization time. This can be unreliable especially in the CI.