zmc
2023-10-12 ed135d79df12a2466b52dae1a82326941211dcc9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
import pytest
 
from pandas.core.dtypes.common import (
    is_bool_dtype,
    is_numeric_dtype,
    is_object_dtype,
    is_string_dtype,
)
 
import pandas as pd
import pandas._testing as tm
from pandas.tests.extension.base.base import BaseExtensionTests
 
 
class BaseGroupbyTests(BaseExtensionTests):
    """Groupby-specific tests."""
 
    def test_grouping_grouper(self, data_for_grouping):
        df = pd.DataFrame(
            {"A": ["B", "B", None, None, "A", "A", "B", "C"], "B": data_for_grouping}
        )
        gr1 = df.groupby("A").grouper.groupings[0]
        gr2 = df.groupby("B").grouper.groupings[0]
 
        tm.assert_numpy_array_equal(gr1.grouping_vector, df.A.values)
        tm.assert_extension_array_equal(gr2.grouping_vector, data_for_grouping)
 
    @pytest.mark.parametrize("as_index", [True, False])
    def test_groupby_extension_agg(self, as_index, data_for_grouping):
        df = pd.DataFrame({"A": [1, 1, 2, 2, 3, 3, 1, 4], "B": data_for_grouping})
        result = df.groupby("B", as_index=as_index).A.mean()
        _, uniques = pd.factorize(data_for_grouping, sort=True)
 
        if as_index:
            index = pd.Index(uniques, name="B")
            expected = pd.Series([3.0, 1.0, 4.0], index=index, name="A")
            self.assert_series_equal(result, expected)
        else:
            expected = pd.DataFrame({"B": uniques, "A": [3.0, 1.0, 4.0]})
            self.assert_frame_equal(result, expected)
 
    def test_groupby_agg_extension(self, data_for_grouping):
        # GH#38980 groupby agg on extension type fails for non-numeric types
        df = pd.DataFrame({"A": [1, 1, 2, 2, 3, 3, 1, 4], "B": data_for_grouping})
 
        expected = df.iloc[[0, 2, 4, 7]]
        expected = expected.set_index("A")
 
        result = df.groupby("A").agg({"B": "first"})
        self.assert_frame_equal(result, expected)
 
        result = df.groupby("A").agg("first")
        self.assert_frame_equal(result, expected)
 
        result = df.groupby("A").first()
        self.assert_frame_equal(result, expected)
 
    def test_groupby_agg_extension_timedelta_cumsum_with_named_aggregation(self):
        # GH#41720
        expected = pd.DataFrame(
            {
                "td": {
                    0: pd.Timedelta("0 days 01:00:00"),
                    1: pd.Timedelta("0 days 01:15:00"),
                    2: pd.Timedelta("0 days 01:15:00"),
                }
            }
        )
        df = pd.DataFrame(
            {
                "td": pd.Series(
                    ["0 days 01:00:00", "0 days 00:15:00", "0 days 01:15:00"],
                    dtype="timedelta64[ns]",
                ),
                "grps": ["a", "a", "b"],
            }
        )
        gb = df.groupby("grps")
        result = gb.agg(td=("td", "cumsum"))
        self.assert_frame_equal(result, expected)
 
    def test_groupby_extension_no_sort(self, data_for_grouping):
        df = pd.DataFrame({"A": [1, 1, 2, 2, 3, 3, 1, 4], "B": data_for_grouping})
        result = df.groupby("B", sort=False).A.mean()
        _, index = pd.factorize(data_for_grouping, sort=False)
 
        index = pd.Index(index, name="B")
        expected = pd.Series([1.0, 3.0, 4.0], index=index, name="A")
        self.assert_series_equal(result, expected)
 
    def test_groupby_extension_transform(self, data_for_grouping):
        valid = data_for_grouping[~data_for_grouping.isna()]
        df = pd.DataFrame({"A": [1, 1, 3, 3, 1, 4], "B": valid})
 
        result = df.groupby("B").A.transform(len)
        expected = pd.Series([3, 3, 2, 2, 3, 1], name="A")
 
        self.assert_series_equal(result, expected)
 
    def test_groupby_extension_apply(self, data_for_grouping, groupby_apply_op):
        df = pd.DataFrame({"A": [1, 1, 2, 2, 3, 3, 1, 4], "B": data_for_grouping})
        df.groupby("B", group_keys=False).apply(groupby_apply_op)
        df.groupby("B", group_keys=False).A.apply(groupby_apply_op)
        df.groupby("A", group_keys=False).apply(groupby_apply_op)
        df.groupby("A", group_keys=False).B.apply(groupby_apply_op)
 
    def test_groupby_apply_identity(self, data_for_grouping):
        df = pd.DataFrame({"A": [1, 1, 2, 2, 3, 3, 1, 4], "B": data_for_grouping})
        result = df.groupby("A").B.apply(lambda x: x.array)
        expected = pd.Series(
            [
                df.B.iloc[[0, 1, 6]].array,
                df.B.iloc[[2, 3]].array,
                df.B.iloc[[4, 5]].array,
                df.B.iloc[[7]].array,
            ],
            index=pd.Index([1, 2, 3, 4], name="A"),
            name="B",
        )
        self.assert_series_equal(result, expected)
 
    def test_in_numeric_groupby(self, data_for_grouping):
        df = pd.DataFrame(
            {
                "A": [1, 1, 2, 2, 3, 3, 1, 4],
                "B": data_for_grouping,
                "C": [1, 1, 1, 1, 1, 1, 1, 1],
            }
        )
 
        dtype = data_for_grouping.dtype
        if (
            is_numeric_dtype(dtype)
            or is_bool_dtype(dtype)
            or dtype.name == "decimal"
            or is_string_dtype(dtype)
            or is_object_dtype(dtype)
            or dtype.kind == "m"  # in particular duration[*][pyarrow]
        ):
            expected = pd.Index(["B", "C"])
            result = df.groupby("A").sum().columns
        else:
            expected = pd.Index(["C"])
            with pytest.raises(TypeError, match="does not support"):
                df.groupby("A").sum().columns
            result = df.groupby("A").sum(numeric_only=True).columns
        tm.assert_index_equal(result, expected)