import numpy as np
import pytest

import pandas as pd
from pandas import (
    Categorical,
    DataFrame,
    NaT,
    Timestamp,
    date_range,
)
import pandas._testing as tm
from pandas.util.version import Version


class TestDataFrameSortValues:
    @pytest.mark.parametrize("dtype", [np.uint8, bool])
    def test_sort_values_sparse_no_warning(self, dtype):
        # GH#45618
        ser = pd.Series(Categorical(["a", "b", "a"], categories=["a", "b", "c"]))
        df = pd.get_dummies(ser, dtype=dtype, sparse=True)

        with tm.assert_produces_warning(None):
            # No warnings about constructing Index from SparseArray
            df.sort_values(by=df.columns.tolist())

    def test_sort_values(self):
        frame = DataFrame(
            [[1, 1, 2], [3, 1, 0], [4, 5, 6]], index=[1, 2, 3], columns=list("ABC")
        )

        # by column (axis=0)
        sorted_df = frame.sort_values(by="A")
        indexer = frame["A"].argsort().values
        expected = frame.loc[frame.index[indexer]]
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by="A", ascending=False)
        indexer = indexer[::-1]
        expected = frame.loc[frame.index[indexer]]
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by="A", ascending=False)
        tm.assert_frame_equal(sorted_df, expected)

        # GH4839
        sorted_df = frame.sort_values(by=["A"], ascending=[False])
        tm.assert_frame_equal(sorted_df, expected)

        # multiple bys
        sorted_df = frame.sort_values(by=["B", "C"])
        expected = frame.loc[[2, 1, 3]]
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by=["B", "C"], ascending=False)
        tm.assert_frame_equal(sorted_df, expected[::-1])

        sorted_df = frame.sort_values(by=["B", "A"], ascending=[True, False])
        tm.assert_frame_equal(sorted_df, expected)

        msg = "No axis named 2 for object type DataFrame"
        with pytest.raises(ValueError, match=msg):
            frame.sort_values(by=["A", "B"], axis=2, inplace=True)

        # by row (axis=1): GH#10806
        sorted_df = frame.sort_values(by=3, axis=1)
        expected = frame
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by=3, axis=1, ascending=False)
        expected = frame.reindex(columns=["C", "B", "A"])
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by=[1, 2], axis="columns")
        expected = frame.reindex(columns=["B", "A", "C"])
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by=[1, 3], axis=1, ascending=[True, False])
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.sort_values(by=[1, 3], axis=1, ascending=False)
        expected = frame.reindex(columns=["C", "B", "A"])
        tm.assert_frame_equal(sorted_df, expected)

        msg = r"Length of ascending \(5\) != length of by \(2\)"
        with pytest.raises(ValueError, match=msg):
            frame.sort_values(by=["A", "B"], axis=0, ascending=[True] * 5)

    def test_sort_values_by_empty_list(self):
        # https://github.com/pandas-dev/pandas/issues/40258
        expected = DataFrame({"a": [1, 4, 2, 5, 3, 6]})
        result = expected.sort_values(by=[])
        tm.assert_frame_equal(result, expected)
        assert result is not expected

    def test_sort_values_inplace(self):
        frame = DataFrame(
            np.random.default_rng(2).standard_normal((4, 4)),
            index=[1, 2, 3, 4],
            columns=["A", "B", "C", "D"],
        )

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(by="A", inplace=True)
        assert return_value is None
        expected = frame.sort_values(by="A")
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(by=1, axis=1, inplace=True)
        assert return_value is None
        expected = frame.sort_values(by=1, axis=1)
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(by="A", ascending=False, inplace=True)
        assert return_value is None
        expected = frame.sort_values(by="A", ascending=False)
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(
            by=["A", "B"], ascending=False, inplace=True
        )
        assert return_value is None
        expected = frame.sort_values(by=["A", "B"], ascending=False)
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_multicolumn(self):
        A = np.arange(5).repeat(20)
        B = np.tile(np.arange(5), 20)
        np.random.default_rng(2).shuffle(A)
        np.random.default_rng(2).shuffle(B)
        frame = DataFrame(
            {"A": A, "B": B, "C": np.random.default_rng(2).standard_normal(100)}
        )

        result = frame.sort_values(by=["A", "B"])
        indexer = np.lexsort((frame["B"], frame["A"]))
        expected = frame.take(indexer)
        tm.assert_frame_equal(result, expected)

        result = frame.sort_values(by=["A", "B"], ascending=False)
        indexer = np.lexsort(
            (frame["B"].rank(ascending=False), frame["A"].rank(ascending=False))
        )
        expected = frame.take(indexer)
        tm.assert_frame_equal(result, expected)

        result = frame.sort_values(by=["B", "A"])
        indexer = np.lexsort((frame["A"], frame["B"]))
        expected = frame.take(indexer)
        tm.assert_frame_equal(result, expected)

    def test_sort_values_multicolumn_uint64(self):
        # GH#9918
        # uint64 multicolumn sort

        df = DataFrame(
            {
                "a": pd.Series([18446637057563306014, 1162265347240853609]),
                "b": pd.Series([1, 2]),
            }
        )
        df["a"] = df["a"].astype(np.uint64)
        result = df.sort_values(["a", "b"])

        expected = DataFrame(
            {
                "a": pd.Series([18446637057563306014, 1162265347240853609]),
                "b": pd.Series([1, 2]),
            },
            index=pd.Index([1, 0]),
        )

        tm.assert_frame_equal(result, expected)

    def test_sort_values_nan(self):
        # GH#3917
        df = DataFrame(
            {"A": [1, 2, np.nan, 1, 6, 8, 4], "B": [9, np.nan, 5, 2, 5, 4, 5]}
        )

        # sort one column only
        expected = DataFrame(
            {"A": [np.nan, 1, 1, 2, 4, 6, 8], "B": [5, 9, 2, np.nan, 5, 5, 4]},
            index=[2, 0, 3, 1, 6, 4, 5],
        )
        sorted_df = df.sort_values(["A"], na_position="first")
        tm.assert_frame_equal(sorted_df, expected)

        expected = DataFrame(
            {"A": [np.nan, 8, 6, 4, 2, 1, 1], "B": [5, 4, 5, 5, np.nan, 9, 2]},
            index=[2, 5, 4, 6, 1, 0, 3],
        )
        sorted_df = df.sort_values(["A"], na_position="first", ascending=False)
        tm.assert_frame_equal(sorted_df, expected)

        expected = df.reindex(columns=["B", "A"])
        sorted_df = df.sort_values(by=1, axis=1, na_position="first")
        tm.assert_frame_equal(sorted_df, expected)

        # na_position='last', order
        expected = DataFrame(
            {"A": [1, 1, 2, 4, 6, 8, np.nan], "B": [2, 9, np.nan, 5, 5, 4, 5]},
            index=[3, 0, 1, 6, 4, 5, 2],
        )
        sorted_df = df.sort_values(["A", "B"])
        tm.assert_frame_equal(sorted_df, expected)

        # na_position='first', order
        expected = DataFrame(
            {"A": [np.nan, 1, 1, 2, 4, 6, 8], "B": [5, 2, 9, np.nan, 5, 5, 4]},
            index=[2, 3, 0, 1, 6, 4, 5],
        )
        sorted_df = df.sort_values(["A", "B"], na_position="first")
        tm.assert_frame_equal(sorted_df, expected)

        # na_position='first', not order
        expected = DataFrame(
            {"A": [np.nan, 1, 1, 2, 4, 6, 8], "B": [5, 9, 2, np.nan, 5, 5, 4]},
            index=[2, 0, 3, 1, 6, 4, 5],
        )
        sorted_df = df.sort_values(["A", "B"], ascending=[1, 0], na_position="first")
        tm.assert_frame_equal(sorted_df, expected)

        # na_position='last', not order
        expected = DataFrame(
            {"A": [8, 6, 4, 2, 1, 1, np.nan], "B": [4, 5, 5, np.nan, 2, 9, 5]},
            index=[5, 4, 6, 1, 3, 0, 2],
        )
        sorted_df = df.sort_values(["A", "B"], ascending=[0, 1], na_position="last")
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_stable_descending_sort(self):
        # GH#6399
        df = DataFrame(
            [[2, "first"], [2, "second"], [1, "a"], [1, "b"]],
            columns=["sort_col", "order"],
        )
        sorted_df = df.sort_values(by="sort_col", kind="mergesort", ascending=False)
        tm.assert_frame_equal(df, sorted_df)

    @pytest.mark.parametrize(
        "expected_idx_non_na, ascending",
        [
            [
                [3, 4, 5, 0, 1, 8, 6, 9, 7, 10, 13, 14],
                [True, True],
            ],
            [
                [0, 3, 4, 5, 1, 8, 6, 7, 10, 13, 14, 9],
                [True, False],
            ],
            [
                [9, 7, 10, 13, 14, 6, 8, 1, 3, 4, 5, 0],
                [False, True],
            ],
            [
                [7, 10, 13, 14, 9, 6, 8, 1, 0, 3, 4, 5],
                [False, False],
            ],
        ],
    )
    @pytest.mark.parametrize("na_position", ["first", "last"])
    def test_sort_values_stable_multicolumn_sort(
        self, expected_idx_non_na, ascending, na_position
    ):
        # GH#38426 Clarify sort_values with mult. columns / labels is stable
        df = DataFrame(
            {
                "A": [1, 2, np.nan, 1, 1, 1, 6, 8, 4, 8, 8, np.nan, np.nan, 8, 8],
                "B": [9, np.nan, 5, 2, 2, 2, 5, 4, 5, 3, 4, np.nan, np.nan, 4, 4],
            }
        )
        # All rows with NaN in col "B" only have unique values in "A", therefore,
        # only the rows with NaNs in "A" have to be treated individually:
        expected_idx = (
            [11, 12, 2] + expected_idx_non_na
            if na_position == "first"
            else expected_idx_non_na + [2, 11, 12]
        )
        expected = df.take(expected_idx)
        sorted_df = df.sort_values(
            ["A", "B"], ascending=ascending, na_position=na_position
        )
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_stable_categorial(self):
        # GH#16793
        df = DataFrame({"x": Categorical(np.repeat([1, 2, 3, 4], 5), ordered=True)})
        expected = df.copy()
        sorted_df = df.sort_values("x", kind="mergesort")
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_datetimes(self):
        # GH#3461, argsort / lexsort differences for a datetime column
        df = DataFrame(
            ["a", "a", "a", "b", "c", "d", "e", "f", "g"],
            columns=["A"],
            index=date_range("20130101", periods=9),
        )
        dts = [
            Timestamp(x)
            for x in [
                "2004-02-11",
                "2004-01-21",
                "2004-01-26",
                "2005-09-20",
                "2010-10-04",
                "2009-05-12",
                "2008-11-12",
                "2010-09-28",
                "2010-09-28",
            ]
        ]
        df["B"] = dts[::2] + dts[1::2]
        df["C"] = 2.0
        df["A1"] = 3.0

        df1 = df.sort_values(by="A")
        df2 = df.sort_values(by=["A"])
        tm.assert_frame_equal(df1, df2)

        df1 = df.sort_values(by="B")
        df2 = df.sort_values(by=["B"])
        tm.assert_frame_equal(df1, df2)

        df1 = df.sort_values(by="B")

        df2 = df.sort_values(by=["C", "B"])
        tm.assert_frame_equal(df1, df2)

    def test_sort_values_frame_column_inplace_sort_exception(
        self, float_frame, using_copy_on_write
    ):
        s = float_frame["A"]
        float_frame_orig = float_frame.copy()
        if using_copy_on_write:
            # INFO(CoW) Series is a new object, so can be changed inplace
            # without modifying original datafame
            s.sort_values(inplace=True)
            tm.assert_series_equal(s, float_frame_orig["A"].sort_values())
            # column in dataframe is not changed
            tm.assert_frame_equal(float_frame, float_frame_orig)
        else:
            with pytest.raises(ValueError, match="This Series is a view"):
                s.sort_values(inplace=True)

        cp = s.copy()
        cp.sort_values()  # it works!

    def test_sort_values_nat_values_in_int_column(self):
        # GH#14922: "sorting with large float and multiple columns incorrect"

        # cause was that the int64 value NaT was considered as "na". Which is
        # only correct for datetime64 columns.

        int_values = (2, int(NaT._value))
        float_values = (2.0, -1.797693e308)

        df = DataFrame(
            {"int": int_values, "float": float_values}, columns=["int", "float"]
        )

        df_reversed = DataFrame(
            {"int": int_values[::-1], "float": float_values[::-1]},
            columns=["int", "float"],
            index=[1, 0],
        )

        # NaT is not a "na" for int64 columns, so na_position must not
        # influence the result:
        df_sorted = df.sort_values(["int", "float"], na_position="last")
        tm.assert_frame_equal(df_sorted, df_reversed)

        df_sorted = df.sort_values(["int", "float"], na_position="first")
        tm.assert_frame_equal(df_sorted, df_reversed)

        # reverse sorting order
        df_sorted = df.sort_values(["int", "float"], ascending=False)
        tm.assert_frame_equal(df_sorted, df)

        # and now check if NaT is still considered as "na" for datetime64
        # columns:
        df = DataFrame(
            {"datetime": [Timestamp("2016-01-01"), NaT], "float": float_values},
            columns=["datetime", "float"],
        )

        df_reversed = DataFrame(
            {"datetime": [NaT, Timestamp("2016-01-01")], "float": float_values[::-1]},
            columns=["datetime", "float"],
            index=[1, 0],
        )

        df_sorted = df.sort_values(["datetime", "float"], na_position="first")
        tm.assert_frame_equal(df_sorted, df_reversed)

        df_sorted = df.sort_values(["datetime", "float"], na_position="last")
        tm.assert_frame_equal(df_sorted, df)

        # Ascending should not affect the results.
        df_sorted = df.sort_values(["datetime", "float"], ascending=False)
        tm.assert_frame_equal(df_sorted, df)

    def test_sort_nat(self):
        # GH 16836

        d1 = [Timestamp(x) for x in ["2016-01-01", "2015-01-01", np.nan, "2016-01-01"]]
        d2 = [
            Timestamp(x)
            for x in ["2017-01-01", "2014-01-01", "2016-01-01", "2015-01-01"]
        ]
        df = DataFrame({"a": d1, "b": d2}, index=[0, 1, 2, 3])

        d3 = [Timestamp(x) for x in ["2015-01-01", "2016-01-01", "2016-01-01", np.nan]]
        d4 = [
            Timestamp(x)
            for x in ["2014-01-01", "2015-01-01", "2017-01-01", "2016-01-01"]
        ]
        expected = DataFrame({"a": d3, "b": d4}, index=[1, 3, 0, 2])
        sorted_df = df.sort_values(by=["a", "b"])
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_na_position_with_categories(self):
        # GH#22556
        # Positioning missing value properly when column is Categorical.
        categories = ["A", "B", "C"]
        category_indices = [0, 2, 4]
        list_of_nans = [np.nan, np.nan]
        na_indices = [1, 3]
        na_position_first = "first"
        na_position_last = "last"
        column_name = "c"

        reversed_categories = sorted(categories, reverse=True)
        reversed_category_indices = sorted(category_indices, reverse=True)
        reversed_na_indices = sorted(na_indices)

        df = DataFrame(
            {
                column_name: Categorical(
                    ["A", np.nan, "B", np.nan, "C"], categories=categories, ordered=True
                )
            }
        )
        # sort ascending with na first
        result = df.sort_values(
            by=column_name, ascending=True, na_position=na_position_first
        )
        expected = DataFrame(
            {
                column_name: Categorical(
                    list_of_nans + categories, categories=categories, ordered=True
                )
            },
            index=na_indices + category_indices,
        )

        tm.assert_frame_equal(result, expected)

        # sort ascending with na last
        result = df.sort_values(
            by=column_name, ascending=True, na_position=na_position_last
        )
        expected = DataFrame(
            {
                column_name: Categorical(
                    categories + list_of_nans, categories=categories, ordered=True
                )
            },
            index=category_indices + na_indices,
        )

        tm.assert_frame_equal(result, expected)

        # sort descending with na first
        result = df.sort_values(
            by=column_name, ascending=False, na_position=na_position_first
        )
        expected = DataFrame(
            {
                column_name: Categorical(
                    list_of_nans + reversed_categories,
                    categories=categories,
                    ordered=True,
                )
            },
            index=reversed_na_indices + reversed_category_indices,
        )

        tm.assert_frame_equal(result, expected)

        # sort descending with na last
        result = df.sort_values(
            by=column_name, ascending=False, na_position=na_position_last
        )
        expected = DataFrame(
            {
                column_name: Categorical(
                    reversed_categories + list_of_nans,
                    categories=categories,
                    ordered=True,
                )
            },
            index=reversed_category_indices + reversed_na_indices,
        )

        tm.assert_frame_equal(result, expected)

    def test_sort_values_nat(self):
        # GH#16836

        d1 = [Timestamp(x) for x in ["2016-01-01", "2015-01-01", np.nan, "2016-01-01"]]
        d2 = [
            Timestamp(x)
            for x in ["2017-01-01", "2014-01-01", "2016-01-01", "2015-01-01"]
        ]
        df = DataFrame({"a": d1, "b": d2}, index=[0, 1, 2, 3])

        d3 = [Timestamp(x) for x in ["2015-01-01", "2016-01-01", "2016-01-01", np.nan]]
        d4 = [
            Timestamp(x)
            for x in ["2014-01-01", "2015-01-01", "2017-01-01", "2016-01-01"]
        ]
        expected = DataFrame({"a": d3, "b": d4}, index=[1, 3, 0, 2])
        sorted_df = df.sort_values(by=["a", "b"])
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_na_position_with_categories_raises(self):
        df = DataFrame(
            {
                "c": Categorical(
                    ["A", np.nan, "B", np.nan, "C"],
                    categories=["A", "B", "C"],
                    ordered=True,
                )
            }
        )

        with pytest.raises(ValueError, match="invalid na_position: bad_position"):
            df.sort_values(by="c", ascending=False, na_position="bad_position")

    @pytest.mark.parametrize("inplace", [True, False])
    @pytest.mark.parametrize(
        "original_dict, sorted_dict, ignore_index, output_index",
        [
            ({"A": [1, 2, 3]}, {"A": [3, 2, 1]}, True, [0, 1, 2]),
            ({"A": [1, 2, 3]}, {"A": [3, 2, 1]}, False, [2, 1, 0]),
            (
                {"A": [1, 2, 3], "B": [2, 3, 4]},
                {"A": [3, 2, 1], "B": [4, 3, 2]},
                True,
                [0, 1, 2],
            ),
            (
                {"A": [1, 2, 3], "B": [2, 3, 4]},
                {"A": [3, 2, 1], "B": [4, 3, 2]},
                False,
                [2, 1, 0],
            ),
        ],
    )
    def test_sort_values_ignore_index(
        self, inplace, original_dict, sorted_dict, ignore_index, output_index
    ):
        # GH 30114
        df = DataFrame(original_dict)
        expected = DataFrame(sorted_dict, index=output_index)
        kwargs = {"ignore_index": ignore_index, "inplace": inplace}

        if inplace:
            result_df = df.copy()
            result_df.sort_values("A", ascending=False, **kwargs)
        else:
            result_df = df.sort_values("A", ascending=False, **kwargs)

        tm.assert_frame_equal(result_df, expected)
        tm.assert_frame_equal(df, DataFrame(original_dict))

    def test_sort_values_nat_na_position_default(self):
        # GH 13230
        expected = DataFrame(
            {
                "A": [1, 2, 3, 4, 4],
                "date": pd.DatetimeIndex(
                    [
                        "2010-01-01 09:00:00",
                        "2010-01-01 09:00:01",
                        "2010-01-01 09:00:02",
                        "2010-01-01 09:00:03",
                        "NaT",
                    ]
                ),
            }
        )
        result = expected.sort_values(["A", "date"])
        tm.assert_frame_equal(result, expected)

    def test_sort_values_item_cache(self, using_array_manager, using_copy_on_write):
        # previous behavior incorrect retained an invalid _item_cache entry
        df = DataFrame(
            np.random.default_rng(2).standard_normal((4, 3)), columns=["A", "B", "C"]
        )
        df["D"] = df["A"] * 2
        ser = df["A"]
        if not using_array_manager:
            assert len(df._mgr.blocks) == 2

        df.sort_values(by="A")

        if using_copy_on_write:
            ser.iloc[0] = 99
            assert df.iloc[0, 0] == df["A"][0]
            assert df.iloc[0, 0] != 99
        else:
            ser.values[0] = 99
            assert df.iloc[0, 0] == df["A"][0]
            assert df.iloc[0, 0] == 99

    def test_sort_values_reshaping(self):
        # GH 39426
        values = list(range(21))
        expected = DataFrame([values], columns=values)
        df = expected.sort_values(expected.index[0], axis=1, ignore_index=True)

        tm.assert_frame_equal(df, expected)

    def test_sort_values_no_by_inplace(self):
        # GH#50643
        df = DataFrame({"a": [1, 2, 3]})
        expected = df.copy()
        result = df.sort_values(by=[], inplace=True)
        tm.assert_frame_equal(df, expected)
        assert result is None

    def test_sort_values_no_op_reset_index(self):
        # GH#52553
        df = DataFrame({"A": [10, 20], "B": [1, 5]}, index=[2, 3])
        result = df.sort_values(by="A", ignore_index=True)
        expected = DataFrame({"A": [10, 20], "B": [1, 5]})
        tm.assert_frame_equal(result, expected)


class TestDataFrameSortKey:  # test key sorting (issue 27237)
    def test_sort_values_inplace_key(self, sort_by_key):
        frame = DataFrame(
            np.random.default_rng(2).standard_normal((4, 4)),
            index=[1, 2, 3, 4],
            columns=["A", "B", "C", "D"],
        )

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(by="A", inplace=True, key=sort_by_key)
        assert return_value is None
        expected = frame.sort_values(by="A", key=sort_by_key)
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(
            by=1, axis=1, inplace=True, key=sort_by_key
        )
        assert return_value is None
        expected = frame.sort_values(by=1, axis=1, key=sort_by_key)
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        return_value = sorted_df.sort_values(
            by="A", ascending=False, inplace=True, key=sort_by_key
        )
        assert return_value is None
        expected = frame.sort_values(by="A", ascending=False, key=sort_by_key)
        tm.assert_frame_equal(sorted_df, expected)

        sorted_df = frame.copy()
        sorted_df.sort_values(
            by=["A", "B"], ascending=False, inplace=True, key=sort_by_key
        )
        expected = frame.sort_values(by=["A", "B"], ascending=False, key=sort_by_key)
        tm.assert_frame_equal(sorted_df, expected)

    def test_sort_values_key(self):
        df = DataFrame(np.array([0, 5, np.nan, 3, 2, np.nan]))

        result = df.sort_values(0)
        expected = df.iloc[[0, 4, 3, 1, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(0, key=lambda x: x + 5)
        expected = df.iloc[[0, 4, 3, 1, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(0, key=lambda x: -x, ascending=False)
        expected = df.iloc[[0, 4, 3, 1, 2, 5]]
        tm.assert_frame_equal(result, expected)

    def test_sort_values_by_key(self):
        df = DataFrame(
            {
                "a": np.array([0, 3, np.nan, 3, 2, np.nan]),
                "b": np.array([0, 2, np.nan, 5, 2, np.nan]),
            }
        )

        result = df.sort_values("a", key=lambda x: -x)
        expected = df.iloc[[1, 3, 4, 0, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(by=["a", "b"], key=lambda x: -x)
        expected = df.iloc[[3, 1, 4, 0, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(by=["a", "b"], key=lambda x: -x, ascending=False)
        expected = df.iloc[[0, 4, 1, 3, 2, 5]]
        tm.assert_frame_equal(result, expected)

    def test_sort_values_by_key_by_name(self):
        df = DataFrame(
            {
                "a": np.array([0, 3, np.nan, 3, 2, np.nan]),
                "b": np.array([0, 2, np.nan, 5, 2, np.nan]),
            }
        )

        def key(col):
            if col.name == "a":
                return -col
            else:
                return col

        result = df.sort_values(by="a", key=key)
        expected = df.iloc[[1, 3, 4, 0, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(by=["a"], key=key)
        expected = df.iloc[[1, 3, 4, 0, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(by="b", key=key)
        expected = df.iloc[[0, 1, 4, 3, 2, 5]]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(by=["a", "b"], key=key)
        expected = df.iloc[[1, 3, 4, 0, 2, 5]]
        tm.assert_frame_equal(result, expected)

    def test_sort_values_key_string(self):
        df = DataFrame(np.array([["hello", "goodbye"], ["hello", "Hello"]]))

        result = df.sort_values(1)
        expected = df[::-1]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values([0, 1], key=lambda col: col.str.lower())
        tm.assert_frame_equal(result, df)

        result = df.sort_values(
            [0, 1], key=lambda col: col.str.lower(), ascending=False
        )
        expected = df.sort_values(1, key=lambda col: col.str.lower(), ascending=False)
        tm.assert_frame_equal(result, expected)

    def test_sort_values_key_empty(self, sort_by_key):
        df = DataFrame(np.array([]))

        df.sort_values(0, key=sort_by_key)
        df.sort_index(key=sort_by_key)

    def test_changes_length_raises(self):
        df = DataFrame({"A": [1, 2, 3]})
        with pytest.raises(ValueError, match="change the shape"):
            df.sort_values("A", key=lambda x: x[:1])

    def test_sort_values_key_axes(self):
        df = DataFrame({0: ["Hello", "goodbye"], 1: [0, 1]})

        result = df.sort_values(0, key=lambda col: col.str.lower())
        expected = df[::-1]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(1, key=lambda col: -col)
        expected = df[::-1]
        tm.assert_frame_equal(result, expected)

    def test_sort_values_key_dict_axis(self):
        df = DataFrame({0: ["Hello", 0], 1: ["goodbye", 1]})

        result = df.sort_values(0, key=lambda col: col.str.lower(), axis=1)
        expected = df.loc[:, ::-1]
        tm.assert_frame_equal(result, expected)

        result = df.sort_values(1, key=lambda col: -col, axis=1)
        expected = df.loc[:, ::-1]
        tm.assert_frame_equal(result, expected)

    @pytest.mark.parametrize("ordered", [True, False])
    def test_sort_values_key_casts_to_categorical(self, ordered):
        # https://github.com/pandas-dev/pandas/issues/36383
        categories = ["c", "b", "a"]
        df = DataFrame({"x": [1, 1, 1], "y": ["a", "b", "c"]})

        def sorter(key):
            if key.name == "y":
                return pd.Series(
                    Categorical(key, categories=categories, ordered=ordered)
                )
            return key

        result = df.sort_values(by=["x", "y"], key=sorter)
        expected = DataFrame(
            {"x": [1, 1, 1], "y": ["c", "b", "a"]}, index=pd.Index([2, 1, 0])
        )

        tm.assert_frame_equal(result, expected)


@pytest.fixture
def df_none():
    return DataFrame(
        {
            "outer": ["a", "a", "a", "b", "b", "b"],
            "inner": [1, 2, 2, 2, 1, 1],
            "A": np.arange(6, 0, -1),
            ("B", 5): ["one", "one", "two", "two", "one", "one"],
        }
    )


@pytest.fixture(params=[["outer"], ["outer", "inner"]])
def df_idx(request, df_none):
    levels = request.param
    return df_none.set_index(levels)


@pytest.fixture(
    params=[
        "inner",  # index level
        ["outer"],  # list of index level
        "A",  # column
        [("B", 5)],  # list of column
        ["inner", "outer"],  # two index levels
        [("B", 5), "outer"],  # index level and column
        ["A", ("B", 5)],  # Two columns
        ["inner", "outer"],  # two index levels and column
    ]
)
def sort_names(request):
    return request.param


@pytest.fixture(params=[True, False])
def ascending(request):
    return request.param


class TestSortValuesLevelAsStr:
    def test_sort_index_level_and_column_label(
        self, df_none, df_idx, sort_names, ascending, request
    ):
        # GH#14353
        if (
            Version(np.__version__) >= Version("1.25")
            and request.node.callspec.id == "df_idx0-inner-True"
        ):
            request.applymarker(
                pytest.mark.xfail(
                    reason=(
                        "pandas default unstable sorting of duplicates"
                        "issue with numpy>=1.25 with AVX instructions"
                    ),
                    strict=False,
                )
            )

        # Get index levels from df_idx
        levels = df_idx.index.names

        # Compute expected by sorting on columns and the setting index
        expected = df_none.sort_values(
            by=sort_names, ascending=ascending, axis=0
        ).set_index(levels)

        # Compute result sorting on mix on columns and index levels
        result = df_idx.sort_values(by=sort_names, ascending=ascending, axis=0)

        tm.assert_frame_equal(result, expected)

    def test_sort_column_level_and_index_label(
        self, df_none, df_idx, sort_names, ascending, request
    ):
        # GH#14353

        # Get levels from df_idx
        levels = df_idx.index.names

        # Compute expected by sorting on axis=0, setting index levels, and then
        # transposing. For some cases this will result in a frame with
        # multiple column levels
        expected = (
            df_none.sort_values(by=sort_names, ascending=ascending, axis=0)
            .set_index(levels)
            .T
        )

        # Compute result by transposing and sorting on axis=1.
        result = df_idx.T.sort_values(by=sort_names, ascending=ascending, axis=1)

        if Version(np.__version__) >= Version("1.25"):
            request.applymarker(
                pytest.mark.xfail(
                    reason=(
                        "pandas default unstable sorting of duplicates"
                        "issue with numpy>=1.25 with AVX instructions"
                    ),
                    strict=False,
                )
            )

        tm.assert_frame_equal(result, expected)

    def test_sort_values_validate_ascending_for_value_error(self):
        # GH41634
        df = DataFrame({"D": [23, 7, 21]})

        msg = 'For argument "ascending" expected type bool, received type str.'
        with pytest.raises(ValueError, match=msg):
            df.sort_values(by="D", ascending="False")

    @pytest.mark.parametrize("ascending", [False, 0, 1, True])
    def test_sort_values_validate_ascending_functional(self, ascending):
        df = DataFrame({"D": [23, 7, 21]})
        indexer = df["D"].argsort().values

        if not ascending:
            indexer = indexer[::-1]

        expected = df.loc[df.index[indexer]]
        result = df.sort_values(by="D", ascending=ascending)
        tm.assert_frame_equal(result, expected)