Test hash aggregate udf with randomly sampled data
(sum_agg_func_fixture)
| 862 | |
| 863 | @pytest.mark.numpy |
| 864 | def test_hash_agg_random(sum_agg_func_fixture): |
| 865 | """Test hash aggregate udf with randomly sampled data""" |
| 866 | |
| 867 | value_num = 1000000 |
| 868 | group_num = 1000 |
| 869 | |
| 870 | arr1 = pa.array(np.repeat(1, value_num), pa.float64()) |
| 871 | arr2 = pa.array(np.random.choice(group_num, value_num), pa.int32()) |
| 872 | |
| 873 | table = pa.table([arr2, arr1], names=['id', 'value']) |
| 874 | |
| 875 | result = table.group_by("id").aggregate([("value", "sum_udf")]) |
| 876 | expected = table.group_by("id").aggregate( |
| 877 | [("value", "sum")]).rename_columns(['id', 'value_sum_udf']) |
| 878 | |
| 879 | assert result.sort_by('id') == expected.sort_by('id') |
| 880 | |
| 881 | |
| 882 | @pytest.mark.pandas |