Function test_table_group_by

python/pyarrow/tests/test_table.py:2870–3023

()

@pytest.mark.acero
def test_table_group_by():
    def sorted_by_keys(d):
        # Ensure a guaranteed order of keys for aggregation results.
        if "keys2" in d:
            keys = tuple(zip(d["keys"], d["keys2"]))
        else:
            keys = d["keys"]
        sorted_keys = sorted(keys)
        sorted_d = {"keys": sorted(d["keys"])}
        for entry in d:
            if entry == "keys":
                continue
            values = dict(zip(keys, d[entry]))
            for k in sorted_keys:
                sorted_d.setdefault(entry, []).append(values[k])
        return sorted_d

    table = pa.table([
        pa.array(["a", "a", "b", "b", "c"]),
        pa.array(["X", "X", "Y", "Z", "Z"]),
        pa.array([1, 2, 3, 4, 5]),
        pa.array([10, 20, 30, 40, 50])
    ], names=["keys", "keys2", "values", "bigvalues"])

    r = table.group_by("keys").aggregate([
        ("values", "hash_sum")
    ])
    assert sorted_by_keys(r.to_pydict()) == {
        "keys": ["a", "b", "c"],
        "values_sum": [3, 7, 5]
    }

    r = table.group_by("keys").aggregate([
        ("values", "hash_sum"),
        ("values", "hash_count")
    ])
    assert sorted_by_keys(r.to_pydict()) == {
        "keys": ["a", "b", "c"],
        "values_sum": [3, 7, 5],
        "values_count": [2, 2, 1]
    }

    # Test without hash_ prefix
    r = table.group_by("keys").aggregate([
        ("values", "sum")
    ])
    assert sorted_by_keys(r.to_pydict()) == {
        "keys": ["a", "b", "c"],
        "values_sum": [3, 7, 5]
    }

    r = table.group_by("keys").aggregate([
        ("values", "max"),
        ("bigvalues", "sum")
    ])
    assert sorted_by_keys(r.to_pydict()) == {
        "keys": ["a", "b", "c"],
        "values_max": [2, 4, 5],