test_frame.py

from __future__ import annotations

from collections import defaultdict
import csv
import datetime
from enum import Enum
import io
import itertools
from pathlib import Path
from typing import (
    TYPE_CHECKING,
    Any,
    Callable,
    Dict,
    Generic,
    Hashable,
    Iterable,
    Iterator,
    List,
    Mapping,
    Tuple,
    TypedDict,
    TypeVar,
    Union,
    cast,
)

import numpy as np
import numpy.typing as npt
import pandas as pd
from pandas._testing import (
    ensure_clean,
    getSeriesData,
)
from pandas.core.resample import Resampler  # noqa: F401
import pytest
from typing_extensions import assert_type
import xarray as xr

from pandas._typing import Scalar

from tests import (
    PD_LTE_20,
    TYPE_CHECKING_INVALID_USAGE,
    check,
)

from pandas.io.formats.style import Styler
from pandas.io.parsers import TextFileReader

DF = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})


def test_types_init() -> None:
    pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]}, index=[2, 1])
    pd.DataFrame(data=[[1, 2, 3], [4, 5, 6]])
    pd.DataFrame(data=itertools.repeat([1, 2, 3], 3))
    pd.DataFrame(data=(range(i) for i in range(5)))
    pd.DataFrame(data=[1, 2, 3, 4], dtype=np.int8)
    pd.DataFrame(
        np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]]),
        columns=["a", "b", "c"],
        dtype=np.int8,
        copy=True,
    )
    check(
        assert_type(pd.DataFrame(0, index=[0, 1], columns=[0, 1]), pd.DataFrame),
        pd.DataFrame,
    )


def test_types_all() -> None:
    df = pd.DataFrame([[False, True], [False, False]], columns=["col1", "col2"])
    check(assert_type(df.all(), "pd.Series[bool]"), pd.Series, np.bool_)
    check(assert_type(df.all(axis=None), bool), np.bool_)


def test_types_any() -> None:
    df = pd.DataFrame([[False, True], [False, False]], columns=["col1", "col2"])
    check(assert_type(df.any(), "pd.Series[bool]"), pd.Series, np.bool_)
    check(assert_type(df.any(axis=None), bool), np.bool_)


def test_types_append() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df2 = pd.DataFrame({"col1": [10, 20], "col2": [30, 40]})
    if TYPE_CHECKING_INVALID_USAGE:
        res1: pd.DataFrame = df.append(df2)  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
        res2: pd.DataFrame = df.append([1, 2, 3])  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
        res3: pd.DataFrame = df.append([[1, 2, 3]])  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
        res4: pd.DataFrame = df.append(  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
            {("a", 1): [1, 2, 3], "b": df2}, ignore_index=True
        )
        res5: pd.DataFrame = df.append(  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
            {1: [1, 2, 3]}, ignore_index=True
        )
        res6: pd.DataFrame = df.append(  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
            {1: [1, 2, 3], "col2": [1, 2, 3]}, ignore_index=True
        )
        res7: pd.DataFrame = df.append(  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
            pd.Series([5, 6]), ignore_index=True
        )
        res8: pd.DataFrame = df.append(  # type: ignore[operator] # pyright: ignore[reportGeneralTypeIssues]
            pd.Series([5, 6], index=["col1", "col2"]), ignore_index=True
        )


def test_types_to_csv() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    csv_df: str = df.to_csv()

    with ensure_clean() as path:
        df.to_csv(path)
        df2: pd.DataFrame = pd.read_csv(path)

    with ensure_clean() as path:
        df.to_csv(Path(path))
        df3: pd.DataFrame = pd.read_csv(Path(path))

    # This keyword was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    with ensure_clean() as path:
        df.to_csv(path, errors="replace")
        df4: pd.DataFrame = pd.read_csv(path)

    # Testing support for binary file handles, added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
    df.to_csv(io.BytesIO(), encoding="utf-8", compression="gzip")

    # Testing support for binary file handles, added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
    df.to_csv(io.BytesIO(), quoting=csv.QUOTE_ALL, encoding="utf-8", compression="gzip")


def test_types_to_csv_when_path_passed() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    with ensure_clean() as file:
        path = Path(file)
        df.to_csv(path)
        df5: pd.DataFrame = pd.read_csv(path)


def test_types_copy() -> None:
    df = pd.DataFrame([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
    df2: pd.DataFrame = df.copy()


def test_types_getitem() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4], 5: [6, 7]})
    i = pd.Index(["col1", "col2"])
    s = pd.Series(["col1", "col2"])
    select_df = pd.DataFrame({"col1": [True, True], "col2": [False, True]})
    a = np.array(["col1", "col2"])
    df["col1"]
    df[5]
    df[["col1", "col2"]]
    df[1:]
    df[s]
    df[a]
    df[select_df]
    df[i]


def test_types_getitem_with_hashable() -> None:
    # Testing getitem support for hashable types that are not scalar
    # Due to the bug in https://github.com/pandas-dev/pandas-stubs/issues/592
    class MyEnum(Enum):
        FIRST = "tayyar"
        SECOND = "haydar"

    df = pd.DataFrame(
        data=[[12.2, 10], [8.8, 15]], columns=[MyEnum.FIRST, MyEnum.SECOND]
    )
    check(assert_type(df[MyEnum.FIRST], pd.Series), pd.Series)
    check(assert_type(df[1:], pd.DataFrame), pd.DataFrame)
    check(assert_type(df[:2], pd.DataFrame), pd.DataFrame)

    df2 = pd.DataFrame(data=[[12.2, 10], [8.8, 15]], columns=[3, 4])
    check(assert_type(df2[3], pd.Series), pd.Series)
    check(assert_type(df2[[3]], pd.DataFrame), pd.DataFrame)
    check(assert_type(df2[[3, 4]], pd.DataFrame), pd.DataFrame)


def test_slice_setitem() -> None:
    # Due to the bug in pandas 1.2.3(https://github.com/pandas-dev/pandas/issues/40440), this is in separate test case
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4], 5: [6, 7]})
    df[1:] = ["a", "b", "c"]


def test_types_setitem() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4], 5: [6, 7]})
    i = pd.Index(["col1", "col2"])
    s = pd.Series(["col1", "col2"])
    a = np.array(["col1", "col2"])
    df["col1"] = [1, 2]
    df[5] = [5, 6]
    df[["col1", "col2"]] = [[1, 2], [3, 4]]
    df[s] = [5, 6]
    df[a] = [[1, 2], [3, 4]]
    df[i] = [8, 9]


def test_types_setitem_mask() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4], 5: [6, 7]})
    select_df = pd.DataFrame({"col1": [True, True], "col2": [False, True]})
    df[select_df] = [1, 2, 3]


def test_types_iloc_iat() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.iloc[1, 1]
    df.iloc[[1], [1]]
    df.iat[0, 0]


def test_types_loc_at() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.loc[[0], "col1"]
    df.at[0, "col1"]
    df.loc[0, "col1"]


def test_types_boolean_indexing() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df[df > 1]
    df[~(df > 1.0)]


def test_types_df_to_df_comparison() -> None:
    df = pd.DataFrame(data={"col1": [1, 2]})
    df2 = pd.DataFrame(data={"col1": [3, 2]})
    res_gt: pd.DataFrame = df > df2
    res_ge: pd.DataFrame = df >= df2
    res_lt: pd.DataFrame = df < df2
    res_le: pd.DataFrame = df <= df2
    res_e: pd.DataFrame = df == df2


def test_types_head_tail() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.head(1)
    df.tail(1)


def test_types_assign() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.assign(col3=lambda frame: frame.sum(axis=1))
    df["col3"] = df.sum(axis=1)


def test_types_sample() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    # GH 67
    check(assert_type(df.sample(frac=0.5), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.sample(n=1), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.sample(n=1, random_state=np.random.default_rng()), pd.DataFrame),
        pd.DataFrame,
    )


def test_types_nlargest_nsmallest() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.nlargest(1, "col1")
    df.nsmallest(1, "col2")


def test_types_filter() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.filter(items=["col1"])
    df.filter(regex="co.*")
    df.filter(like="1")


def test_types_setting() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df["col1"] = 1
    df[df == 1] = 7


def test_types_drop() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    check(assert_type(df.drop("col1", axis=1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(columns=["col1"]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop([0]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(index=[0]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(columns=["col1"]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(index=1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(labels=0), pd.DataFrame), pd.DataFrame)
    assert assert_type(df.drop([0, 0], inplace=True), None) is None
    to_drop: list[str] = ["col1"]
    check(assert_type(df.drop(columns=to_drop), pd.DataFrame), pd.DataFrame)
    # GH 302
    check(assert_type(df.drop(pd.Index([1])), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(index=pd.Index([1])), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.drop(columns=pd.Index(["col1"])), pd.DataFrame), pd.DataFrame)


def test_types_dropna() -> None:
    df = pd.DataFrame(data={"col1": [np.nan, np.nan], "col2": [3, np.nan]})
    res: pd.DataFrame = df.dropna()
    res2: pd.DataFrame = df.dropna(axis=1, thresh=1)
    res3: None = df.dropna(axis=0, how="all", subset=["col1"], inplace=True)


def test_types_fillna() -> None:
    df = pd.DataFrame(data={"col1": [np.nan, np.nan], "col2": [3, np.nan]})
    res: pd.DataFrame = df.fillna(0)
    res2: None = df.fillna(method="pad", axis=1, inplace=True)


def test_types_sort_index() -> None:
    df = pd.DataFrame(data={"col1": [1, 2, 3, 4]}, index=[5, 1, 3, 2])
    df2 = pd.DataFrame(data={"col1": [1, 2, 3, 4]}, index=["a", "b", "c", "d"])
    res: pd.DataFrame = df.sort_index()
    level1 = (1, 2)
    res2: pd.DataFrame = df.sort_index(ascending=False, level=level1)
    level2: list[str] = ["a", "b", "c"]
    res3: pd.DataFrame = df2.sort_index(level=level2)
    res4: pd.DataFrame = df.sort_index(ascending=False, level=3)
    res5: None = df.sort_index(kind="mergesort", inplace=True)


# This was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
def test_types_sort_index_with_key() -> None:
    df = pd.DataFrame(data={"col1": [1, 2, 3, 4]}, index=["a", "b", "C", "d"])
    res: pd.DataFrame = df.sort_index(key=lambda k: k.str.lower())


def test_types_set_index() -> None:
    df = pd.DataFrame(
        data={"col1": [1, 2, 3, 4], "col2": ["a", "b", "c", "d"]}, index=[5, 1, 3, 2]
    )
    res: pd.DataFrame = df.set_index("col1")
    res2: pd.DataFrame = df.set_index("col1", drop=False)
    res3: pd.DataFrame = df.set_index("col1", append=True)
    res4: pd.DataFrame = df.set_index("col1", verify_integrity=True)
    res5: pd.DataFrame = df.set_index(["col1", "col2"])
    res6: None = df.set_index("col1", inplace=True)
    # GH 140
    res7: pd.DataFrame = df.set_index(pd.Index(["w", "x", "y", "z"]))


def test_types_query() -> None:
    df = pd.DataFrame(data={"col1": [1, 2, 3, 4], "col2": [3, 0, 1, 7]})
    res: pd.DataFrame = df.query("col1 > col2")
    res2: None = df.query("col1 % col2 == 0", inplace=True)


def test_types_eval() -> None:
    df = pd.DataFrame(data={"col1": [1, 2, 3, 4], "col2": [3, 0, 1, 7]})
    df.eval("col1 > col2")
    res: None = df.eval("C = col1 % col2 == 0", inplace=True)


def test_types_sort_values() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    res: pd.DataFrame = df.sort_values("col1")
    res2: None = df.sort_values("col1", ascending=False, inplace=True)
    res3: pd.DataFrame = df.sort_values(by=["col1", "col2"], ascending=[True, False])


# This was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
def test_types_sort_values_with_key() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    res: pd.DataFrame = df.sort_values(by="col1", key=lambda k: -k)


def test_types_shift() -> None:
    df = pd.DataFrame(data={"col1": [1, 1], "col2": [3, 4]})
    df.shift()
    df.shift(1)
    df.shift(-1)


def test_types_rank() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.rank(axis=0, na_option="bottom")
    df.rank(method="min", pct=True)
    df.rank(method="dense", ascending=True)
    df.rank(method="first", numeric_only=True)


def test_types_mean() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    s1: pd.Series = df.mean()
    s2: pd.Series = df.mean(axis=0)
    df2: pd.DataFrame = df.groupby(level=0).mean()
    if PD_LTE_20:
        df3: pd.DataFrame = df.groupby(axis=1, level=0).mean()
        df4: pd.DataFrame = df.groupby(axis=1, level=0, dropna=True).mean()
    s3: pd.Series = df.mean(axis=1, skipna=True, numeric_only=False)


def test_types_median() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    s1: pd.Series = df.median()
    s2: pd.Series = df.median(axis=0)
    df2: pd.DataFrame = df.groupby(level=0).median()
    if PD_LTE_20:
        df3: pd.DataFrame = df.groupby(axis=1, level=0).median()
        df4: pd.DataFrame = df.groupby(axis=1, level=0, dropna=True).median()
    s3: pd.Series = df.median(axis=1, skipna=True, numeric_only=False)


def test_types_itertuples() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    res1: Iterable[tuple[Any, ...]] = df.itertuples()
    res2: Iterable[tuple[Any, ...]] = df.itertuples(index=False, name="Foobar")
    res3: Iterable[tuple[Any, ...]] = df.itertuples(index=False, name=None)


def test_types_sum() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.sum()
    df.sum(axis=1)


def test_types_cumsum() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.cumsum()
    df.sum(axis=0)


def test_types_min() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.min()
    df.min(axis=0)


def test_types_max() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.max()
    df.max(axis=0)


def test_types_quantile() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.quantile([0.25, 0.5])
    df.quantile(0.75)
    df.quantile()
    # GH 81
    df.quantile(np.array([0.25, 0.75]))


def test_types_clip() -> None:
    df = pd.DataFrame(data={"col1": [20, 12], "col2": [3, 14]})
    df.clip(lower=5, upper=15)


def test_types_abs() -> None:
    df = pd.DataFrame(data={"col1": [-5, 1], "col2": [3, -14]})
    df.abs()


def test_types_var() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [1, 4]})
    df.var()
    df.var(axis=1, ddof=1)
    df.var(skipna=True, numeric_only=False)


def test_types_std() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [1, 4]})
    df.std()
    df.std(axis=1, ddof=1)
    df.std(skipna=True, numeric_only=False)


def test_types_idxmin() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.idxmin()
    df.idxmin(axis=0)


def test_types_idxmax() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.idxmax()
    df.idxmax(axis=0)


# This was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
def test_types_value_counts() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [1, 4]})
    df.value_counts()


def test_types_unique() -> None:
    # This is really more for of a Series test
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [1, 4]})
    df["col1"].unique()


def test_types_apply() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4], "col3": [5, 6]})

    def returns_scalar(x: pd.Series) -> int:
        return 2

    def returns_series(x: pd.Series) -> pd.Series:
        return x**2

    def returns_listlike_of_2(x: pd.Series) -> tuple[int, int]:
        return (7, 8)

    def returns_listlike_of_3(x: pd.Series) -> tuple[int, int, int]:
        return (7, 8, 9)

    def returns_dict(x: pd.Series) -> dict[str, int]:
        return {"col4": 7, "col5": 8}

    # Misc checks
    check(assert_type(df.apply(np.exp), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.apply(str), "pd.Series[str]"), pd.Series, str)

    # GH 393
    def gethead(s: pd.Series, y: int) -> pd.Series:
        return s.head(y)

    check(assert_type(df.apply(gethead, args=(4,)), pd.DataFrame), pd.DataFrame)

    # Check various return types for default result_type (None) with default axis (0)
    check(assert_type(df.apply(returns_scalar), "pd.Series[int]"), pd.Series, np.int64)
    check(assert_type(df.apply(returns_series), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.apply(returns_listlike_of_3), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.apply(returns_dict), pd.Series), pd.Series)

    # Check various return types for result_type="expand" with default axis (0)
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "expand" to a scalar return
        assert_type(df.apply(returns_scalar, result_type="expand"), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(df.apply(returns_series, result_type="expand"), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(
            df.apply(returns_listlike_of_3, result_type="expand"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(df.apply(returns_dict, result_type="expand"), pd.DataFrame),
        pd.DataFrame,
    )

    # Check various return types for result_type="reduce" with default axis (0)
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "reduce" to a scalar return
        assert_type(df.apply(returns_scalar, result_type="reduce"), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "reduce" to a series return
        assert_type(df.apply(returns_series, result_type="reduce"), pd.Series),
        pd.Series,  # This technically returns a pd.Series[pd.Series], but typing does not support that
    )
    check(
        assert_type(df.apply(returns_listlike_of_3, result_type="reduce"), pd.Series),
        pd.Series,
    )
    check(
        assert_type(df.apply(returns_dict, result_type="reduce"), pd.Series), pd.Series
    )

    # Check various return types for default result_type (None) with axis=1
    check(
        assert_type(df.apply(returns_scalar, axis=1), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(assert_type(df.apply(returns_series, axis=1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.apply(returns_listlike_of_3, axis=1), pd.Series), pd.Series)
    check(assert_type(df.apply(returns_dict, axis=1), pd.Series), pd.Series)

    # Check various return types for result_type="expand" with axis=1
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "expand" to a scalar return
        assert_type(
            df.apply(returns_scalar, axis=1, result_type="expand"), "pd.Series[int]"
        ),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(
            df.apply(returns_series, axis=1, result_type="expand"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            df.apply(returns_listlike_of_3, axis=1, result_type="expand"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(df.apply(returns_dict, axis=1, result_type="expand"), pd.DataFrame),
        pd.DataFrame,
    )

    # Check various return types for result_type="reduce" with axis=1
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "reduce" to a scalar return
        assert_type(
            df.apply(returns_scalar, axis=1, result_type="reduce"), "pd.Series[int]"
        ),
        pd.Series,
        np.int64,
    )
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "reduce" to a series return
        assert_type(
            df.apply(returns_series, axis=1, result_type="reduce"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            df.apply(returns_listlike_of_3, axis=1, result_type="reduce"), pd.Series
        ),
        pd.Series,
    )
    check(
        assert_type(df.apply(returns_dict, axis=1, result_type="reduce"), pd.Series),
        pd.Series,
    )

    # Check various return types for result_type="broadcast" with axis=0 and axis=1
    check(
        # Note that technically it does not make sense
        # to pass a result_type of "broadcast" to a scalar return
        assert_type(df.apply(returns_scalar, result_type="broadcast"), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.apply(returns_series, result_type="broadcast"), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        # Can only broadcast a list-like of 2 elements, not 3, because there are 2 rows
        assert_type(
            df.apply(returns_listlike_of_2, result_type="broadcast"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        # Note that technicaly it does not make sense
        # to pass a result_type of "broadcast" to a scalar return
        assert_type(
            df.apply(returns_scalar, axis=1, result_type="broadcast"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            df.apply(returns_series, axis=1, result_type="broadcast"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            # Can only broadcast a list-like of 3 elements, not 2,
            # as there are 3 columns
            df.apply(returns_listlike_of_3, axis=1, result_type="broadcast"),
            pd.DataFrame,
        ),
        pd.DataFrame,
    )
    # Since dicts will be assigned to elements of np.ndarray inside broadcasting,
    # we need to use a DataFrame with object dtype to make the assignment possible.
    df2 = pd.DataFrame({"col1": ["a", "b"], "col2": ["c", "d"]})
    check(
        assert_type(df2.apply(returns_dict, result_type="broadcast"), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(
            df2.apply(returns_dict, axis=1, result_type="broadcast"), pd.DataFrame
        ),
        pd.DataFrame,
    )

    # Test various other positional/keyword argument combinations
    # to ensure all overloads are supported
    check(
        assert_type(df.apply(returns_scalar, axis=0), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(
            df.apply(returns_scalar, axis=0, result_type=None), "pd.Series[int]"
        ),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(df.apply(returns_scalar, 0, False, None), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(
            df.apply(returns_scalar, 0, False, result_type=None), "pd.Series[int]"
        ),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(
            df.apply(returns_scalar, 0, raw=False, result_type=None), "pd.Series[int]"
        ),
        pd.Series,
        np.int64,
    )


def test_types_applymap() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df.applymap(lambda x: x**2)
    df.applymap(np.exp)
    df.applymap(str)
    # na_action parameter was added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
    df.applymap(np.exp, na_action="ignore")
    df.applymap(str, na_action=None)


def test_types_element_wise_arithmetic() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})
    df2 = pd.DataFrame(data={"col1": [10, 20], "col3": [3, 4]})

    res_add1: pd.DataFrame = df + df2
    res_add2: pd.DataFrame = df.add(df2, fill_value=0)

    res_sub: pd.DataFrame = df - df2
    res_sub2: pd.DataFrame = df.sub(df2, fill_value=0)

    res_mul: pd.DataFrame = df * df2
    res_mul2: pd.DataFrame = df.mul(df2, fill_value=0)

    res_div: pd.DataFrame = df / df2
    res_div2: pd.DataFrame = df.div(df2, fill_value=0)

    res_floordiv: pd.DataFrame = df // df2
    res_floordiv2: pd.DataFrame = df.floordiv(df2, fill_value=0)

    res_mod: pd.DataFrame = df % df2
    res_mod2: pd.DataFrame = df.mod(df2, fill_value=0)

    res_pow: pd.DataFrame = df2**df
    res_pow2: pd.DataFrame = df2.pow(df, fill_value=0)

    # divmod operation was added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
    # noinspection PyTypeChecker
    res_divmod: tuple[pd.DataFrame, pd.DataFrame] = divmod(df, df2)
    res_divmod2: tuple[pd.DataFrame, pd.DataFrame] = df.__divmod__(df2)
    res_rdivmod: tuple[pd.DataFrame, pd.DataFrame] = df.__rdivmod__(df2)


def test_types_scalar_arithmetic() -> None:
    df = pd.DataFrame(data={"col1": [2, 1], "col2": [3, 4]})

    res_add1: pd.DataFrame = df + 1
    res_add2: pd.DataFrame = df.add(1, fill_value=0)

    res_sub: pd.DataFrame = df - 1
    res_sub2: pd.DataFrame = df.sub(1, fill_value=0)

    res_mul: pd.DataFrame = df * 2
    res_mul2: pd.DataFrame = df.mul(2, fill_value=0)

    res_div: pd.DataFrame = df / 2
    res_div2: pd.DataFrame = df.div(2, fill_value=0)

    res_floordiv: pd.DataFrame = df // 2
    res_floordiv2: pd.DataFrame = df.floordiv(2, fill_value=0)

    res_mod: pd.DataFrame = df % 2
    res_mod2: pd.DataFrame = df.mod(2, fill_value=0)

    res_pow: pd.DataFrame = df**2
    res_pow1: pd.DataFrame = df**0
    res_pow2: pd.DataFrame = df**0.213
    res_pow3: pd.DataFrame = df.pow(0.5)


def test_types_melt() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    df.melt()
    df.melt(id_vars=["col1"], value_vars=["col2"])
    df.melt(
        id_vars=["col1"],
        value_vars=["col2"],
        var_name="someVariable",
        value_name="someValue",
    )

    pd.melt(df)
    pd.melt(df, id_vars=["col1"], value_vars=["col2"])
    pd.melt(
        df,
        id_vars=["col1"],
        value_vars=["col2"],
        var_name="someVariable",
        value_name="someValue",
    )


def test_types_pivot() -> None:
    df = pd.DataFrame(
        data={
            "col1": ["first", "second", "third", "fourth"],
            "col2": [50, 70, 56, 111],
            "col3": ["A", "B", "B", "A"],
            "col4": [100, 102, 500, 600],
        }
    )
    check(
        assert_type(
            df.pivot(index="col1", columns="col3", values="col2"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(df.pivot(index="col1", columns="col3"), pd.DataFrame), pd.DataFrame
    )
    check(
        assert_type(
            df.pivot(index="col1", columns="col3", values=["col2", "col4"]),
            pd.DataFrame,
        ),
        pd.DataFrame,
    )
    check(assert_type(df.pivot(columns="col3"), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.pivot(columns="col3", values="col2"), pd.DataFrame), pd.DataFrame
    )


def test_types_groupby() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5], "col3": [0, 1, 0]})
    df.index.name = "ind"
    df.groupby(by="col1")
    df.groupby(level="ind")
    df.groupby(by="col1", sort=False, as_index=True)
    df.groupby(by=["col1", "col2"])
    # GH 284
    df.groupby(df["col1"] > 2)
    df.groupby([df["col1"] > 2, df["col2"] % 2 == 1])
    df.groupby(lambda x: x)
    df.groupby([lambda x: x % 2, lambda x: x % 3])
    df.groupby(np.array([1, 0, 1]))
    df.groupby([np.array([1, 0, 0]), np.array([0, 0, 1])])
    df.groupby({1: 1, 2: 2, 3: 3})
    df.groupby([{1: 1, 2: 1, 3: 2}, {1: 1, 2: 2, 3: 2}])
    df.groupby(df.index)
    df.groupby([pd.Index([1, 0, 0]), pd.Index([0, 0, 1])])
    df.groupby(pd.Grouper(level=0))
    df.groupby([pd.Grouper(level=0), pd.Grouper(key="col1")])

    df1: pd.DataFrame = df.groupby(by="col1").agg("sum")
    df2: pd.DataFrame = df.groupby(level="ind").aggregate("sum")
    df3: pd.DataFrame = df.groupby(by="col1", sort=False, as_index=True).transform(
        lambda x: x.max()
    )
    df4: pd.DataFrame = df.groupby(by=["col1", "col2"]).count()
    df5: pd.DataFrame = df.groupby(by=["col1", "col2"]).filter(lambda x: x["col1"] > 0)
    df6: pd.DataFrame = df.groupby(by=["col1", "col2"]).nunique()
    df7: pd.DataFrame = df.groupby(by="col1").apply(sum)
    df8: pd.DataFrame = df.groupby("col1").transform("sum")
    s1: pd.Series = df.set_index("col1")["col2"]
    s2: pd.Series = s1.groupby("col1").transform("sum")


def test_types_groupby_methods() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5], "col3": [0, 1, 0]})
    check(assert_type(df.groupby("col1").sum(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby("col1").prod(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby("col1").sample(), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.groupby("col1").value_counts(normalize=False), "pd.Series[int]"),
        pd.Series,
        np.int64,
    )
    check(
        assert_type(
            df.groupby("col1").value_counts(subset=None, normalize=True),
            "pd.Series[float]",
        ),
        pd.Series,
        float,
    )
    check(assert_type(df.groupby("col1").idxmax(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby("col1").idxmin(), pd.DataFrame), pd.DataFrame)


def test_types_groupby_agg() -> None:
    df = pd.DataFrame(
        data={"col1": [1, 1, 2], "col2": [3, 4, 5], "col3": [0, 1, 0], 0: [-1, -1, -1]}
    )
    check(assert_type(df.groupby("col1")["col3"].agg(min), pd.Series), pd.Series)
    check(
        assert_type(df.groupby("col1")["col3"].agg([min, max]), pd.DataFrame),
        pd.DataFrame,
    )
    check(assert_type(df.groupby("col1").agg("min"), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby("col1").agg(min), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.groupby("col1").agg(["min", "max"]), pd.DataFrame), pd.DataFrame
    )
    check(assert_type(df.groupby("col1").agg([min, max]), pd.DataFrame), pd.DataFrame)
    agg_dict1 = {"col2": "min", "col3": "max", 0: "sum"}
    check(assert_type(df.groupby("col1").agg(agg_dict1), pd.DataFrame), pd.DataFrame)
    agg_dict2 = {"col2": min, "col3": max, 0: min}
    check(assert_type(df.groupby("col1").agg(agg_dict2), pd.DataFrame), pd.DataFrame)

    def wrapped_min(x: Any) -> Any:
        return x.min()

    # Here, MyPy infers dict[object, object], so it must be explicitly annotated
    agg_dict3: dict[str | int, str | Callable] = {
        "col2": min,
        "col3": "max",
        0: wrapped_min,
    }
    check(assert_type(df.groupby("col1").agg(agg_dict3), pd.DataFrame), pd.DataFrame)
    agg_dict4 = {"col2": "sum"}
    check(assert_type(df.groupby("col1").agg(agg_dict4), pd.DataFrame), pd.DataFrame)
    agg_dict5 = {0: "sum"}
    check(assert_type(df.groupby("col1").agg(agg_dict5), pd.DataFrame), pd.DataFrame)
    named_agg = pd.NamedAgg(column="col2", aggfunc="max")
    check(
        assert_type(df.groupby("col1").agg(new_col=named_agg), pd.DataFrame),
        pd.DataFrame,
    )
    # GH#187
    cols: list[str] = ["col1", "col2"]
    check(assert_type(df.groupby(by=cols).sum(), pd.DataFrame), pd.DataFrame)

    cols_opt: list[str | None] = ["col1", "col2"]
    check(assert_type(df.groupby(by=cols_opt).sum(), pd.DataFrame), pd.DataFrame)

    cols_mixed: list[str | int] = ["col1", 0]
    check(assert_type(df.groupby(by=cols_mixed).sum(), pd.DataFrame), pd.DataFrame)


# This was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
def test_types_group_by_with_dropna_keyword() -> None:
    df = pd.DataFrame(
        data={"col1": [1, 1, 2, 1], "col2": [2, None, 1, 2], "col3": [3, 4, 3, 2]}
    )
    df.groupby(by="col2", dropna=True).sum()
    df.groupby(by="col2", dropna=False).sum()
    df.groupby(by="col2").sum()


def test_types_groupby_any() -> None:
    df = pd.DataFrame(
        data={
            "col1": [1, 1, 2],
            "col2": [True, False, False],
            "col3": [False, False, False],
        }
    )
    check(assert_type(df.groupby("col1").any(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby("col1").all(), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.groupby("col1")["col2"].any(), "pd.Series[bool]"),
        pd.Series,
        np.bool_,
    )
    check(
        assert_type(df.groupby("col1")["col2"].any(), "pd.Series[bool]"),
        pd.Series,
        np.bool_,
    )


def test_types_merge() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    df2 = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [0, 1, 0]})
    df.merge(df2)
    df.merge(df2, on="col1")
    df.merge(df2, on="col1", how="left")
    df.merge(df2, on=["col1", "col2"], how="left")
    df.merge(df2, on=("col1", "col2"), how="left")
    df.merge(df2, on=("col1", "col2"), how="left", suffixes=(None, "s"))
    df.merge(df2, on=("col1", "col2"), how="left", suffixes=("t", "s"))
    df.merge(df2, on=("col1", "col2"), how="left", suffixes=("a", None))
    df.merge(df2, how="cross")  # GH 289
    columns = ["col1", "col2"]
    df.merge(df2, on=columns)


def test_types_plot() -> None:
    if TYPE_CHECKING:  # skip pytest
        df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
        df.plot.hist()
        df.plot.scatter(x="col2", y="col1")


def test_types_window() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    df.expanding()
    if PD_LTE_20:
        df.expanding(axis=1)
        df.rolling(2, axis=1, center=True)
    if TYPE_CHECKING_INVALID_USAGE:
        df.expanding(axis=1, center=True)  # type: ignore[call-arg] # pyright: ignore[reportGeneralTypeIssues]

    df.rolling(2)

    check(
        assert_type(df.rolling(2).agg("max"), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg(max), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg(["max", "min"]), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg([max, min]), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg({"col2": "max"}), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg({"col2": max}), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg({"col2": ["max", "min"]}), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.rolling(2).agg({"col2": [max, min]}), pd.DataFrame),
        pd.DataFrame,
    )


def test_types_cov() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    df.cov()
    df.cov(min_periods=1)
    # ddof param was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    df.cov(ddof=2)


def test_types_to_numpy() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    check(assert_type(df.to_numpy(), np.ndarray), np.ndarray)
    check(assert_type(df.to_numpy(dtype="str", copy=True), np.ndarray), np.ndarray)
    # na_value param was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    check(assert_type(df.to_numpy(na_value=0), np.ndarray), np.ndarray)


def test_to_markdown() -> None:
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    check(assert_type(df.to_markdown(), str), str)
    check(assert_type(df.to_markdown(None), str), str)
    check(assert_type(df.to_markdown(buf=None, mode="wt"), str), str)
    # index param was added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    check(assert_type(df.to_markdown(index=False), str), str)
    with ensure_clean() as path:
        check(assert_type(df.to_markdown(path), None), type(None))
    with ensure_clean() as path:
        check(assert_type(df.to_markdown(Path(path)), None), type(None))
    sio = io.StringIO()
    check(assert_type(df.to_markdown(sio), None), type(None))


def test_types_to_feather() -> None:
    pytest.importorskip("pyarrow")
    df = pd.DataFrame(data={"col1": [1, 1, 2], "col2": [3, 4, 5]})
    with ensure_clean() as path:
        df.to_feather(path)
        # kwargs for pyarrow.feather.write_feather added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
        df.to_feather(path, compression="zstd", compression_level=3, chunksize=2)

        # to_feather has been able to accept a buffer since pandas 1.0.0
        # See https://pandas.pydata.org/docs/whatsnew/v1.0.0.html
        # Docstring and type were updated in 1.2.0.
        # https://github.com/pandas-dev/pandas/pull/35408
        with open(path, mode="wb") as file:
            df.to_feather(file)


# compare() method added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
def test_types_compare() -> None:
    df1 = pd.DataFrame(
        data={"col1": [1, 1, 2, 1], "col2": [2, None, 1, 2], "col3": [3, 4, 3, 2]}
    )
    df2 = pd.DataFrame(
        data={"col1": [1, 2, 5, 6], "col2": [3, 4, 1, 1], "col3": [3, 4, 3, 2]}
    )
    df1.compare(df2)
    df2.compare(df1, align_axis=0, keep_shape=True, keep_equal=True)


def test_types_agg() -> None:
    df = pd.DataFrame([[1, 2, 3], [4, 5, 6], [7, 8, 9]], columns=["A", "B", "C"])
    check(assert_type(df.agg("min"), pd.Series), pd.Series)
    check(assert_type(df.agg(min), pd.Series), pd.Series)
    check(assert_type(df.agg(["min", "max"]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.agg([min, max]), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.agg({"A": ["min", "max"], "B": "min"}), pd.DataFrame),
        pd.DataFrame,
    )
    check(assert_type(df.agg({"A": [min, max], "B": min}), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(
            df.agg(x=("A", max), y=("B", "min"), z=("C", np.mean)), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(assert_type(df.agg("mean", axis=1), pd.Series), pd.Series)


def test_types_aggregate() -> None:
    df = pd.DataFrame([[1, 2, 3], [4, 5, 6], [7, 8, 9]], columns=["A", "B", "C"])
    check(assert_type(df.aggregate("min"), pd.Series), pd.Series)
    check(assert_type(df.aggregate(min), pd.Series), pd.Series)
    check(assert_type(df.aggregate(["min", "max"]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.aggregate([min, max]), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.aggregate({"A": ["min", "max"], "B": "min"}), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.aggregate({"A": [min, max], "B": min}), pd.DataFrame),
        pd.DataFrame,
    )


def test_types_transform() -> None:
    df = pd.DataFrame([[1, 2, 3], [4, 5, 6], [7, 8, 9]], columns=["A", "B", "C"])
    check(assert_type(df.transform("abs"), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.transform(abs), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.transform(["abs", "sqrt"]), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.transform([abs, np.sqrt]), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.transform({"A": ["abs", "sqrt"], "B": "abs"}), pd.DataFrame),
        pd.DataFrame,
    )
    check(
        assert_type(df.transform({"A": [abs, np.sqrt], "B": abs}), pd.DataFrame),
        pd.DataFrame,
    )


def test_types_describe() -> None:
    df = pd.DataFrame(
        data={
            "col1": [1, 2, -4],
            "col2": [
                np.datetime64("2000-01-01"),
                np.datetime64("2010-01-01"),
                np.datetime64("2010-01-01"),
            ],
        }
    )
    df.describe()
    df.describe(percentiles=[0.5], include="all")
    df.describe(exclude=[np.number])


def test_types_to_string() -> None:
    df = pd.DataFrame(
        data={
            "col1": [1, None, -4],
            "col2": [
                np.datetime64("2000-01-01"),
                np.datetime64("2010-01-01"),
                np.datetime64("2010-01-01"),
            ],
        }
    )
    df.to_string(
        index=True,
        col_space=2,
        header=["a", "b"],
        na_rep="0",
        justify="left",
        max_rows=2,
        min_rows=0,
        max_cols=2,
        show_dimensions=True,
        line_width=3,
    )
    # col_space accepting list or dict added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    df.to_string(col_space=[1, 2])
    df.to_string(col_space={"col1": 1, "col2": 3})


def test_types_to_html() -> None:
    df = pd.DataFrame(
        data={
            "col1": [1, None, -4],
            "col2": [
                np.datetime64("2000-01-01"),
                np.datetime64("2010-01-01"),
                np.datetime64("2010-01-01"),
            ],
        }
    )
    df.to_html(
        index=True,
        col_space=2,
        header=True,
        na_rep="0",
        justify="left",
        max_rows=2,
        max_cols=2,
        show_dimensions=True,
    )
    # col_space accepting list or dict added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    df.to_html(col_space=[1, 2])
    df.to_html(col_space={"col1": 1, "col2": 3})


def test_types_resample() -> None:
    df = pd.DataFrame({"values": [2, 11, 3, 13, 14, 18, 17, 19]})
    df["date"] = pd.date_range("01/01/2018", periods=8, freq="W")
    df.resample("M", on="date")
    # origin and offset params added in 1.1.0 https://pandas.pydata.org/docs/whatsnew/v1.1.0.html
    df.resample("20min", origin="epoch", offset=pd.Timedelta(2, "minutes"), on="date")


def test_types_to_dict() -> None:
    data = pd.DataFrame({"a": [1], "b": [2]})
    data.to_dict(orient="records")
    data.to_dict(orient="dict")
    data.to_dict(orient="list")
    data.to_dict(orient="series")
    data.to_dict(orient="split")
    data.to_dict(orient="index")

    # orient param accepting "tight" added in 1.4.0 https://pandas.pydata.org/docs/whatsnew/v1.4.0.html
    data.to_dict(orient="tight")


def test_types_from_dict() -> None:
    pd.DataFrame.from_dict({"col_1": [3, 2, 1, 0], "col_2": ["a", "b", "c", "d"]})
    pd.DataFrame.from_dict({1: [3, 2, 1, 0], 2: ["a", "b", "c", "d"]})
    pd.DataFrame.from_dict({"a": {1: 2}, "b": {3: 4, 1: 4}}, orient="index")
    pd.DataFrame.from_dict({"a": {"row1": 2}, "b": {"row2": 4, "row1": 4}})
    pd.DataFrame.from_dict({"a": (1, 2, 3), "b": (2, 4, 5)})
    pd.DataFrame.from_dict(
        data={"col_1": {"a": 1}, "col_2": {"a": 1, "b": 2}}, orient="columns"
    )
    # orient param accepting "tight" added in 1.4.0 https://pandas.pydata.org/docs/whatsnew/v1.4.0.html
    pd.DataFrame.from_dict(
        data={
            "index": [("a", "b"), ("a", "c")],
            "columns": [("x", 1), ("y", 2)],
            "data": [[1, 3], [2, 4]],
            "index_names": ["n1", "n2"],
            "column_names": ["z1", "z2"],
        },
        orient="tight",
    )


def test_pipe() -> None:
    def foo(df: pd.DataFrame) -> pd.DataFrame:
        return pd.DataFrame(df)

    val = (
        pd.DataFrame(
            {
                "price": [10, 11, 9, 13, 14, 18, 17, 19],
                "volume": [50, 60, 40, 100, 50, 100, 40, 50],
            }
        )
        .assign(week_starting=pd.date_range("01/01/2018", periods=8, freq="W"))
        .resample("M", on="week_starting")
        .pipe(foo)
    )

    check(assert_type(val, pd.DataFrame), pd.DataFrame)

    check(assert_type(pd.DataFrame({"a": [1]}).pipe(foo), pd.DataFrame), pd.DataFrame)

    def bar(val: Styler) -> Styler:
        return val

    check(
        assert_type(pd.DataFrame({"a": [1], "b": [1]}).style.pipe(bar), Styler), Styler
    )

    def baz(val: Styler) -> str:
        return val.to_latex()

    check(assert_type(pd.DataFrame({"a": [1], "b": [1]}).style.pipe(baz), str), str)


# set_flags() method added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
def test_types_set_flags() -> None:
    pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"]).set_flags(
        allows_duplicate_labels=False
    )
    pd.DataFrame([[1, 2], [8, 9]], columns=["A", "A"]).set_flags(
        allows_duplicate_labels=True
    )
    pd.DataFrame([[1, 2], [8, 9]], columns=["A", "A"])


def test_types_to_parquet() -> None:
    pytest.importorskip("pyarrow")
    pytest.importorskip("fastparquet")
    df = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"]).set_flags(
        allows_duplicate_labels=False
    )
    with ensure_clean() as path:
        df.to_parquet(Path(path))
        # to_parquet() returns bytes when no path given since 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
        b: bytes = df.to_parquet()


def test_types_to_latex() -> None:
    df = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"])
    df.to_latex(
        columns=["A"], label="some_label", caption="some_caption", multirow=True
    )
    df.to_latex(escape=False, decimal=",", column_format="r")
    # position param was added in 1.2.0 https://pandas.pydata.org/docs/whatsnew/v1.2.0.html
    df.to_latex(position="some")
    df.to_latex(caption=("cap1", "cap2"))


def test_types_explode() -> None:
    df = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"])
    res1: pd.DataFrame = df.explode("A")
    res2: pd.DataFrame = df.explode("A", ignore_index=False)
    res3: pd.DataFrame = df.explode("A", ignore_index=True)
    res4: pd.DataFrame = df.explode(["A", "B"])


def test_types_rename() -> None:
    df = pd.DataFrame(columns=["a"])
    col_map = {"a": "b"}
    df.rename(columns=col_map)
    df.rename(columns={"a": "b"})
    df.rename(columns={1: "b"})
    # Apparently all of these calls are accepted by pandas
    df.rename(columns={None: "b"})
    df.rename(columns={"": "b"})
    df.rename(columns={(2, 1): "b"})
    df.rename(columns=lambda s: s.upper())


def test_types_eq() -> None:
    df1 = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"])
    res1: pd.DataFrame = df1 == 1
    df2 = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"])
    res2: pd.DataFrame = df1 == df2


def test_types_as_type() -> None:
    df1 = pd.DataFrame([[1, 2], [8, 9]], columns=["A", "B"])
    df2: pd.DataFrame = df1.astype({"A": "int32"})


def test_types_dot() -> None:
    df1 = pd.DataFrame([[0, 1, -2, -1], [1, 1, 1, 1]])
    df2 = pd.DataFrame([[0, 1], [1, 2], [-1, -1], [2, 0]])
    s1 = pd.Series([1, 1, 2, 1])
    np_array = np.array([[0, 1], [1, 2], [-1, -1], [2, 0]])
    df3: pd.DataFrame = df1 @ df2
    df4: pd.DataFrame = df1.dot(df2)
    df5: pd.DataFrame = df1 @ np_array
    df6: pd.DataFrame = df1.dot(np_array)
    df7: pd.Series = df1 @ s1
    df8: pd.Series = df1.dot(s1)


def test_types_regressions() -> None:
    # https://github.com/microsoft/python-type-stubs/issues/32
    df = pd.DataFrame({"x": [1.0, 2.0, 3.0], "y": [4.0, 5, 6]})
    df2: pd.DataFrame = df.astype(int)

    # https://github.com/microsoft/python-type-stubs/issues/38
    df0: pd.DataFrame = pd.DataFrame({"x": [12, 34], "y": [78, 9]})
    ds: pd.DataFrame = df.sort_values(["x", "y"], ascending=[True, False])

    # https://github.com/microsoft/python-type-stubs/issues/55
    df3 = pd.DataFrame([["a", 1], ["b", 2]], columns=["let", "num"]).set_index("let")
    df4: pd.DataFrame = df3.reset_index()
    df5: pd.DataFrame = df4[["num"]]

    # https://github.com/microsoft/python-type-stubs/issues/58
    df1 = pd.DataFrame(columns=["a", "b", "c"])
    df2 = pd.DataFrame(columns=["a", "c"])
    df6: pd.DataFrame = df1.drop(columns=df2.columns)

    # https://github.com/microsoft/python-type-stubs/issues/60
    df1 = pd.DataFrame([["a", 1], ["b", 2]], columns=["let", "num"]).set_index("let")
    s2 = df1["num"]
    res: pd.DataFrame = pd.merge(s2, df1, left_index=True, right_index=True)

    # https://github.com/microsoft/python-type-stubs/issues/62
    df7: pd.DataFrame = pd.DataFrame({"x": [1, 2, 3]}, index=pd.Index(["a", "b", "c"]))
    index: pd.Index = pd.Index(["b"])
    df8: pd.DataFrame = df7.loc[index]

    # https://github.com/microsoft/python-type-stubs/issues/31
    df = pd.DataFrame({"A": [1, 2, 3], "B": [5, 6, 7]})
    column1: pd.DataFrame = df.iloc[:, [0]]
    column2: pd.Series = df.iloc[:, 0]

    df = pd.DataFrame(
        {
            "a_col": list(range(10)),
            "a_nother": list(range(10)),
            "b_col": list(range(10)),
        }
    )
    df.loc[:, lambda df: df.columns.str.startswith("a_")]

    df = df[::-1]

    # https://github.com/microsoft/python-type-stubs/issues/69
    s1 = pd.Series([1, 2, 3])
    s2 = pd.Series([4, 5, 6])
    df = pd.concat([s1, s2], axis=1)
    ss1: pd.Series = pd.concat([s1, s2], axis=0)
    ss2: pd.Series = pd.concat([s1, s2])

    # https://github.com/microsoft/python-type-stubs/issues/110
    d: datetime.date = pd.Timestamp("2021-01-01")
    tslist: list[pd.Timestamp] = list(pd.to_datetime(["2022-01-01", "2022-01-02"]))
    sseries: pd.Series = pd.Series(tslist)
    sseries_plus1: pd.Series = sseries + pd.Timedelta(1, "d")

    # https://github.com/microsoft/pylance-release/issues/2133
    dr = pd.date_range(start="2021-12-01", periods=24, freq="H")
    time = dr.strftime("%H:%M:%S")

    # https://github.com/microsoft/python-type-stubs/issues/115
    df = pd.DataFrame({"A": [1, 2, 3], "B": [5, 6, 7]})
    pd.DatetimeIndex(
        data=df["A"],
        tz=None,
        normalize=False,
        closed=None,
        ambiguous="NaT",
        copy=True,
    )


def test_read_csv() -> None:
    with ensure_clean() as path:
        Path(path).write_text("A,B\n1,2")
        check(assert_type(pd.read_csv(path), pd.DataFrame), pd.DataFrame)
        check(
            assert_type(pd.read_csv(path, iterator=False), pd.DataFrame), pd.DataFrame
        )
        check(
            assert_type(
                pd.read_csv(path, iterator=False, chunksize=None), pd.DataFrame
            ),
            pd.DataFrame,
        )

        with check(
            assert_type(pd.read_csv(path, chunksize=1), TextFileReader), TextFileReader
        ):
            pass
        with check(
            assert_type(pd.read_csv(path, iterator=False, chunksize=1), TextFileReader),
            TextFileReader,
        ):
            pass
        with check(
            assert_type(pd.read_csv(path, iterator=True), TextFileReader),
            TextFileReader,
        ):
            pass
        with check(
            assert_type(
                pd.read_csv(path, iterator=True, chunksize=None), TextFileReader
            ),
            TextFileReader,
        ):
            pass
        with check(
            assert_type(pd.read_csv(path, iterator=True, chunksize=1), TextFileReader),
            TextFileReader,
        ):
            pass

        # https://github.com/microsoft/python-type-stubs/issues/118
        check(
            assert_type(pd.read_csv(path, storage_options=None), pd.DataFrame),
            pd.DataFrame,
        )

        # Allow a variety of dict types for the converters parameter
        converters1 = {"A": str, "B": str}
        check(
            assert_type(pd.read_csv(path, converters=converters1), pd.DataFrame),
            pd.DataFrame,
        )
        converters2 = {"A": str, "B": float}
        check(
            assert_type(pd.read_csv(path, converters=converters2), pd.DataFrame),
            pd.DataFrame,
        )
        converters3 = {0: str, 1: str}
        check(
            assert_type(pd.read_csv(path, converters=converters3), pd.DataFrame),
            pd.DataFrame,
        )
        converters4 = {0: str, 1: float}
        check(
            assert_type(pd.read_csv(path, converters=converters4), pd.DataFrame),
            pd.DataFrame,
        )
        converters5: dict[int | str, Callable[[str], Any]] = {
            0: str,
            "A": float,
        }
        check(
            assert_type(pd.read_csv(path, converters=converters5), pd.DataFrame),
            pd.DataFrame,
        )

        class ReadCsvKwargs(TypedDict):
            converters: dict[int, Callable[[str], Any]]

        read_csv_kwargs: ReadCsvKwargs = {"converters": {0: int}}

        check(
            assert_type(pd.read_csv(path, **read_csv_kwargs), pd.DataFrame),
            pd.DataFrame,
        )

        # Check value covariance for various other parameters too (these only accept a str key)
        na_values = {"A": ["1"], "B": ["1"]}
        check(
            assert_type(pd.read_csv(path, na_values=na_values), pd.DataFrame),
            pd.DataFrame,
        )

    # There are several possible inputs for parse_dates
    with ensure_clean() as path:
        Path(path).write_text("Date,Year,Month,Day\n20221125,2022,11,25")
        parse_dates_1 = ["Date"]
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_1), pd.DataFrame),
            pd.DataFrame,
        )
        check(
            assert_type(
                pd.read_csv(path, index_col="Date", parse_dates=True), pd.DataFrame
            ),
            pd.DataFrame,
        )
        parse_dates_2 = {"combined_date": ["Year", "Month", "Day"]}
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_2), pd.DataFrame),
            pd.DataFrame,
        )
        parse_dates_3 = {"combined_date": [1, 2, 3]}
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_3), pd.DataFrame),
            pd.DataFrame,
        )
        # MyPy calls this Dict[str, object] by default which necessitates the explicit annotation (Pyright does not)
        parse_dates_4: dict[str, list[str | int]] = {"combined_date": [1, "Month", 3]}
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_4), pd.DataFrame),
            pd.DataFrame,
        )
        parse_dates_5 = [0]
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_5), pd.DataFrame),
            pd.DataFrame,
        )
        parse_dates_6 = [[1, 2, 3]]
        check(
            assert_type(pd.read_csv(path, parse_dates=parse_dates_6), pd.DataFrame),
            pd.DataFrame,
        )


def test_groupby_series_methods() -> None:
    df = pd.DataFrame({"x": [1, 2, 2, 3, 3], "y": [10, 20, 30, 40, 50]})
    gb = df.groupby("x")["y"]
    check(assert_type(gb.describe(), pd.DataFrame), pd.DataFrame)
    gb.count().loc[2]
    gb.pct_change().loc[2]
    gb.bfill().loc[2]
    gb.cummax().loc[2]
    gb.cummin().loc[2]
    gb.cumprod().loc[2]
    gb.cumsum().loc[2]
    gb.ffill().loc[2]
    gb.first().loc[2]
    gb.head().loc[2]
    gb.last().loc[2]
    gb.max().loc[2]
    gb.mean().loc[2]
    gb.median().loc[2]
    gb.min().loc[2]
    gb.nlargest().loc[2]
    gb.nsmallest().loc[2]
    gb.nth(0).loc[1]


def test_indexslice_setitem():
    df = pd.DataFrame(
        {"x": [1, 2, 2, 3], "y": [1, 2, 3, 4], "z": [10, 20, 30, 40]}
    ).set_index(["x", "y"])
    s = pd.Series([-1, -2])
    df.loc[pd.IndexSlice[2, :]] = s.values
    df.loc[pd.IndexSlice[2, :], "z"] = [200, 300]
    # GH 314
    df.loc[pd.IndexSlice[pd.Index([2, 3]), :], "z"] = 99


def test_indexslice_getitem():
    # GH 300
    df = (
        pd.DataFrame({"x": [1, 2, 2, 3, 4], "y": [10, 20, 30, 40, 10]})
        .assign(z=lambda df: df.x * df.y)
        .set_index(["x", "y"])
    )
    ind = pd.Index([2, 3])
    check(assert_type(pd.IndexSlice[ind, :], "tuple[pd.Index, slice]"), tuple)
    check(assert_type(df.loc[pd.IndexSlice[ind, :]], pd.DataFrame), pd.DataFrame)
    check(assert_type(df.loc[pd.IndexSlice[1:2]], pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.loc[pd.IndexSlice[:, df["z"] > 40], :], pd.DataFrame),
        pd.DataFrame,
    )
    check(assert_type(df.loc[pd.IndexSlice[2, 30], "z"], Scalar), np.int64)
    check(
        assert_type(df.loc[pd.IndexSlice[[2, 4], [20, 40]], :], pd.DataFrame),
        pd.DataFrame,
    )
    # GH 314
    check(
        assert_type(df.loc[pd.IndexSlice[pd.Index([2, 4]), :], "z"], pd.Series),
        pd.Series,
    )


def test_compute_values():
    df = pd.DataFrame({"x": [1, 2, 3, 4]})
    s: pd.Series = pd.Series([10, 20, 30, 40])
    result: pd.Series = df["x"] + s.values


# https://github.com/microsoft/python-type-stubs/issues/164
def test_sum_get_add() -> None:
    df = pd.DataFrame({"x": [1, 2, 3, 4, 5], "y": [10, 20, 30, 40, 50]})
    s = df["x"]
    check(assert_type(s, pd.Series), pd.Series)
    summer = df.sum(axis=1)
    check(assert_type(summer, pd.Series), pd.Series)

    s2: pd.Series = s + summer
    s3: pd.Series = s + df["y"]
    s4: pd.Series = summer + summer


def test_getset_untyped() -> None:
    result: int = 10
    df = pd.DataFrame({"x": [1, 2, 3, 4, 5], "y": [10, 20, 30, 40, 50]})
    # Tests that Dataframe.__getitem__ needs to return untyped series.
    result = df["x"].max()


def test_getmultiindex_columns() -> None:
    mi = pd.MultiIndex.from_product([[1, 2], ["a", "b"]])
    df = pd.DataFrame([[1, 2, 3, 4], [10, 20, 30, 40]], columns=mi)
    li: list[tuple[int, str]] = [(1, "a"), (2, "b")]
    res1: pd.DataFrame = df[[(1, "a"), (2, "b")]]
    res2: pd.DataFrame = df[li]
    res3: pd.DataFrame = df[
        [(i, s) for i in [1] for s in df.columns.get_level_values(1)]
    ]
    res4: pd.DataFrame = df[[df.columns[0]]]
    check(assert_type(df[df.columns[0]], pd.Series), pd.Series)
    check(assert_type(df[li[0]], pd.Series), pd.Series)


def test_frame_getitem_isin() -> None:
    df = pd.DataFrame({"x": [1, 2, 3, 4, 5]}, index=[1, 2, 3, 4, 5])
    check(assert_type(df[df.index.isin([1, 3, 5])], pd.DataFrame), pd.DataFrame)


def test_to_excel() -> None:
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})

    with ensure_clean() as path:
        df.to_excel(path, engine="openpyxl")
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)
    with ensure_clean() as path:
        df.to_excel(Path(path), engine="openpyxl")
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)
    with ensure_clean() as path:
        df.to_excel(path, engine="openpyxl", startrow=1, startcol=1, header=False)
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)
    with ensure_clean() as path:
        df.to_excel(path, engine="openpyxl", sheet_name="sheet", index=False)
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)
    with ensure_clean() as path:
        df.to_excel(path, engine="openpyxl", header=["x", "y"])
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)
    with ensure_clean() as path:
        df.to_excel(path, engine="openpyxl", columns=["col1"])
        check(assert_type(pd.read_excel(path), pd.DataFrame), pd.DataFrame)


def test_join() -> None:
    float_frame = pd.DataFrame(getSeriesData())
    # GH 29
    left = float_frame["A"].to_frame()
    seriesB = float_frame["B"]
    frameCD = float_frame[["C", "D"]]
    right: list[pd.Series | pd.DataFrame] = [seriesB, frameCD]
    check(assert_type(left.join(right), pd.DataFrame), pd.DataFrame)
    check(assert_type(left.join(right, validate="1:1"), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(left.join(right, validate="one_to_one"), pd.DataFrame), pd.DataFrame
    )
    check(assert_type(left.join(right, validate="1:m"), pd.DataFrame), pd.DataFrame)


def test_types_ffill() -> None:
    # GH 44
    df = pd.DataFrame([[1, 2, 3]])
    check(assert_type(df.ffill(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.ffill(inplace=False), pd.DataFrame), pd.DataFrame)
    assert assert_type(df.ffill(inplace=True), None) is None


def test_types_bfill() -> None:
    # GH 44
    df = pd.DataFrame([[1, 2, 3]])
    check(assert_type(df.bfill(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.bfill(inplace=False), pd.DataFrame), pd.DataFrame)
    assert assert_type(df.bfill(inplace=True), None) is None


def test_types_replace() -> None:
    # GH 44
    df = pd.DataFrame([[1, 2, 3]])
    check(assert_type(df.replace(1, 2), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.replace(1, 2, inplace=False), pd.DataFrame), pd.DataFrame)
    assert assert_type(df.replace(1, 2, inplace=True), None) is None


def test_loop_dataframe() -> None:
    # GH 70
    df = pd.DataFrame({"x": [1, 2, 3]})
    for c in df:
        check(assert_type(df[c], pd.Series), pd.Series)


def test_groupby_index() -> None:
    # GH 42
    df = pd.DataFrame(
        data={"col1": [1, 1, 2], "col2": [3, 4, 5], "col3": [0, 1, 0]}
    ).set_index("col1")
    check(assert_type(df.groupby(df.index).min(), pd.DataFrame), pd.DataFrame)


def test_iloc_npint() -> None:
    # GH 69
    df = pd.DataFrame({"a": [10, 20, 30], "b": [20, 40, 60], "c": [30, 60, 90]})
    iloc = np.argmin(np.random.standard_normal(3))
    df.iloc[iloc]


# https://github.com/pandas-dev/pandas-stubs/issues/143
def test_iloc_tuple() -> None:
    df = pd.DataFrame({"Char": ["A", "B", "C"], "Number": [1, 2, 3]})
    df = df.iloc[0:2,]


def test_set_columns() -> None:
    # GH 73
    df = pd.DataFrame({"a": [1, 2, 3], "b": [0.0, 1, 1]})
    # Next lines should work, but it is a mypy bug
    # https://github.com/python/mypy/issues/3004
    # pyright accepts this, so we only type check for pyright,
    # and also test the code with pytest
    df.columns = ["c", "d"]  # type: ignore[assignment]
    df.columns = [1, 2]  # type: ignore[assignment]
    df.columns = [1, "a"]  # type: ignore[assignment]
    df.columns = np.array([1, 2])  # type: ignore[assignment]
    df.columns = pd.Series([1, 2])  # type: ignore[assignment]
    df.columns = np.array([1, "a"])  # type: ignore[assignment]
    df.columns = pd.Series([1, "a"])  # type: ignore[assignment]
    df.columns = (1, 2)  # type: ignore[assignment]
    df.columns = (1, "a")  # type: ignore[assignment]
    if TYPE_CHECKING_INVALID_USAGE:
        df.columns = "abc"  # type: ignore[assignment] # pyright: ignore[reportGeneralTypeIssues]


def test_frame_index_numpy() -> None:
    # GH 80
    i = np.array([1.0, 2.0])
    pd.DataFrame([[1.0, 2.0], [3.0, 4.0]], columns=["a", "b"], index=i)


def test_frame_stack() -> None:
    multicol2 = pd.MultiIndex.from_tuples([("weight", "kg"), ("height", "m")])
    df_multi_level_cols2 = pd.DataFrame(
        [[1.0, 2.0], [3.0, 4.0]], index=["cat", "dog"], columns=multicol2
    )

    check(
        assert_type(
            df_multi_level_cols2.stack(0), Union[pd.DataFrame, "pd.Series[Any]"]
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            df_multi_level_cols2.stack([0, 1]), Union[pd.DataFrame, "pd.Series[Any]"]
        ),
        pd.Series,
    )


def test_frame_reindex() -> None:
    # GH 84
    df = pd.DataFrame({"a": [1, 2, 3]}, index=[0, 1, 2])
    df.reindex([2, 1, 0])


def test_frame_ndarray_assignmment() -> None:
    # GH 100
    df_a = pd.DataFrame({"a": [0.0] * 10})
    df_a.iloc[:, :] = np.array([[-1.0]] * 10)

    df_b = pd.DataFrame({"a": [0.0] * 10, "b": [1.0] * 10})
    df_b.iloc[:, :] = np.array([[-1.0, np.inf]] * 10)


def test_not_hashable() -> None:
    # GH 113
    assert assert_type(pd.DataFrame.__hash__, None) is None
    assert assert_type(pd.DataFrame().__hash__, None) is None
    assert assert_type(pd.Series.__hash__, None) is None
    assert assert_type(pd.Series([], dtype=object).__hash__, None) is None
    assert assert_type(pd.Index.__hash__, None) is None
    assert assert_type(pd.Index([]).__hash__, None) is None

    def test_func(h: Hashable):
        pass

    if TYPE_CHECKING_INVALID_USAGE:
        test_func(pd.DataFrame())  # type: ignore[arg-type] # pyright: ignore[reportGeneralTypeIssues]
        test_func(pd.Series([], dtype=object))  # type: ignore[arg-type] # pyright: ignore[reportGeneralTypeIssues]
        test_func(pd.Index([]))  # type: ignore[arg-type] # pyright: ignore[reportGeneralTypeIssues]


def test_columns_mixlist() -> None:
    # GH 97
    df = pd.DataFrame({"a": [1, 2, 3], 1: [3, 4, 5]})
    key: list[int | str]
    key = [1]
    check(assert_type(df[key], pd.DataFrame), pd.DataFrame)


def test_frame_scalars_slice() -> None:
    # GH 133
    # scalars:
    # str, bytes, datetime.date, datetime.datetime, datetime.timedelta, bool, int,
    # float, complex, Timestamp, Timedelta

    str_ = "a"
    bytes_ = b"7"
    date = datetime.date(1999, 12, 31)
    datetime_ = datetime.datetime(1999, 12, 31)
    timedelta = datetime.datetime(2000, 1, 1) - datetime.datetime(1999, 12, 31)
    bool_ = True
    int_ = 2
    float_ = 3.14
    complex_ = 1.0 + 3.0j
    timestamp = pd.Timestamp(0)
    pd_timedelta = pd.Timedelta(0, unit="D")
    none = None
    idx = [
        str_,
        bytes_,
        date,
        datetime_,
        timedelta,
        bool_,
        int_,
        float_,
        complex_,
        timestamp,
        pd_timedelta,
        none,
    ]
    values = np.arange(len(idx))[:, None] + np.arange(len(idx))
    df = pd.DataFrame(values, columns=idx, index=idx)

    # Note: bool_ cannot be tested since the index is object and pandas does not
    # support boolean access using loc except when the index is boolean
    check(assert_type(df.loc[str_], pd.Series), pd.Series)
    check(assert_type(df.loc[bytes_], pd.Series), pd.Series)
    check(assert_type(df.loc[date], pd.Series), pd.Series)
    check(assert_type(df.loc[datetime_], pd.Series), pd.Series)
    check(assert_type(df.loc[timedelta], pd.Series), pd.Series)
    check(assert_type(df.loc[int_], pd.Series), pd.Series)
    check(assert_type(df.loc[float_], pd.Series), pd.Series)
    check(assert_type(df.loc[complex_], pd.Series), pd.Series)
    check(assert_type(df.loc[timestamp], pd.Series), pd.Series)
    check(assert_type(df.loc[pd_timedelta], pd.Series), pd.Series)
    check(assert_type(df.loc[none], pd.Series), pd.Series)

    check(assert_type(df.loc[:, str_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, bytes_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, date], pd.Series), pd.Series)
    check(assert_type(df.loc[:, datetime_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, timedelta], pd.Series), pd.Series)
    check(assert_type(df.loc[:, int_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, float_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, complex_], pd.Series), pd.Series)
    check(assert_type(df.loc[:, timestamp], pd.Series), pd.Series)
    check(assert_type(df.loc[:, pd_timedelta], pd.Series), pd.Series)
    check(assert_type(df.loc[:, none], pd.Series), pd.Series)


def test_boolean_loc() -> None:
    # Booleans can only be used in loc when the index is boolean
    df = pd.DataFrame([[0, 1], [1, 0]], columns=[True, False], index=[True, False])
    check(assert_type(df.loc[True], pd.Series), pd.Series)
    check(assert_type(df.loc[:, False], pd.Series), pd.Series)


def test_groupby_result() -> None:
    # GH 142
    df = pd.DataFrame({"a": [0, 1, 2], "b": [4, 5, 6], "c": [7, 8, 9]})
    iterator = df.groupby(["a", "b"]).__iter__()
    assert_type(iterator, Iterator[Tuple[Tuple, pd.DataFrame]])
    index, value = next(iterator)
    assert_type((index, value), Tuple[Tuple, pd.DataFrame])

    check(assert_type(index, Tuple), tuple, np.int64)
    check(assert_type(value, pd.DataFrame), pd.DataFrame)

    iterator2 = df.groupby("a").__iter__()
    assert_type(iterator2, Iterator[Tuple[Scalar, pd.DataFrame]])
    index2, value2 = next(iterator2)
    assert_type((index2, value2), Tuple[Scalar, pd.DataFrame])

    check(assert_type(index2, Scalar), int)
    check(assert_type(value2, pd.DataFrame), pd.DataFrame)

    # GH 674
    # grouping by pd.MultiIndex should always resolve to a tuple as well
    multi_index = pd.MultiIndex.from_frame(df[["a", "b"]])
    iterator3 = df.groupby(multi_index).__iter__()
    assert_type(iterator3, Iterator[Tuple[Tuple, pd.DataFrame]])
    index3, value3 = next(iterator3)
    assert_type((index3, value3), Tuple[Tuple, pd.DataFrame])

    check(assert_type(index3, Tuple), tuple, int)
    check(assert_type(value3, pd.DataFrame), pd.DataFrame)

    # Want to make sure these cases are differentiated
    for (k1, k2), g in df.groupby(["a", "b"]):
        pass

    for kk, g in df.groupby("a"):
        pass

    for (k1, k2), g in df.groupby(multi_index):
        pass


def test_groupby_result_for_scalar_indexes() -> None:
    # GH 674
    dates = pd.date_range("2020-01-01", "2020-12-31")
    df = pd.DataFrame({"date": dates, "days": 1})
    period_index = pd.PeriodIndex(df.date, freq="M")
    iterator = df.groupby(period_index).__iter__()
    assert_type(iterator, Iterator[Tuple[pd.Period, pd.DataFrame]])
    index, value = next(iterator)
    assert_type((index, value), Tuple[pd.Period, pd.DataFrame])

    check(assert_type(index, pd.Period), pd.Period)
    check(assert_type(value, pd.DataFrame), pd.DataFrame)

    dt_index = pd.DatetimeIndex(dates)
    iterator2 = df.groupby(dt_index).__iter__()
    assert_type(iterator2, Iterator[Tuple[pd.Timestamp, pd.DataFrame]])
    index2, value2 = next(iterator2)
    assert_type((index2, value2), Tuple[pd.Timestamp, pd.DataFrame])

    check(assert_type(index2, pd.Timestamp), pd.Timestamp)
    check(assert_type(value2, pd.DataFrame), pd.DataFrame)

    tdelta_index = pd.TimedeltaIndex(dates - pd.Timestamp("2020-01-01"))
    iterator3 = df.groupby(tdelta_index).__iter__()
    assert_type(iterator3, Iterator[Tuple[pd.Timedelta, pd.DataFrame]])
    index3, value3 = next(iterator3)
    assert_type((index3, value3), Tuple[pd.Timedelta, pd.DataFrame])

    check(assert_type(index3, pd.Timedelta), pd.Timedelta)
    check(assert_type(value3, pd.DataFrame), pd.DataFrame)

    intervals: list[pd.Interval[pd.Timestamp]] = [
        pd.Interval(date, date + pd.DateOffset(days=1), closed="left") for date in dates
    ]
    interval_index = pd.IntervalIndex(intervals)
    assert_type(interval_index, "pd.IntervalIndex[pd.Interval[pd.Timestamp]]")
    iterator4 = df.groupby(interval_index).__iter__()
    assert_type(iterator4, Iterator[Tuple["pd.Interval[pd.Timestamp]", pd.DataFrame]])
    index4, value4 = next(iterator4)
    assert_type((index4, value4), Tuple["pd.Interval[pd.Timestamp]", pd.DataFrame])

    check(assert_type(index4, "pd.Interval[pd.Timestamp]"), pd.Interval)
    check(assert_type(value4, pd.DataFrame), pd.DataFrame)

    for p, g in df.groupby(period_index):
        pass

    for dt, g in df.groupby(dt_index):
        pass

    for tdelta, g in df.groupby(tdelta_index):
        pass

    for interval, g in df.groupby(interval_index):
        pass


def test_groupby_result_for_ambiguous_indexes() -> None:
    # GH 674
    df = pd.DataFrame({"a": [0, 1, 2], "b": [4, 5, 6], "c": [7, 8, 9]})
    # this will use pd.Index which is ambiguous
    iterator = df.groupby(df.index).__iter__()
    assert_type(iterator, Iterator[Tuple[Any, pd.DataFrame]])
    index, value = next(iterator)
    assert_type((index, value), Tuple[Any, pd.DataFrame])

    check(assert_type(index, Any), int)
    check(assert_type(value, pd.DataFrame), pd.DataFrame)

    # categorical indexes are also ambiguous
    categorical_index = pd.CategoricalIndex(df.a)
    iterator2 = df.groupby(categorical_index).__iter__()
    assert_type(iterator2, Iterator[Tuple[Any, pd.DataFrame]])
    index2, value2 = next(iterator2)
    assert_type((index2, value2), Tuple[Any, pd.DataFrame])

    check(assert_type(index2, Any), int)
    check(assert_type(value2, pd.DataFrame), pd.DataFrame)


def test_setitem_list():
    # GH 153
    lst1: list[str] = ["a", "b", "c"]
    lst2: list[int] = [1, 2, 3]
    lst3: list[float] = [4.0, 5.0, 6.0]
    lst4: list[tuple[str, int]] = [("a", 1), ("b", 2), ("c", 3)]
    lst5: list[complex] = [0 + 1j, 0 + 2j, 0 + 3j]

    columns: list[Hashable] = [
        "a",
        "b",
        "c",
        1,
        2,
        3,
        4.0,
        5.0,
        6.0,
        ("a", 1),
        ("b", 2),
        ("c", 3),
        0 + 1j,
        0 + 2j,
        0 + 3j,
    ]

    df = pd.DataFrame(np.empty((3, 15)), columns=columns)

    check(assert_type(df.set_index(lst1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.set_index(lst2), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.set_index(lst3), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.set_index(lst4), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.set_index(lst5), pd.DataFrame), pd.DataFrame)

    iter1: Iterator[str] = (v for v in lst1)
    iter2: Iterator[tuple[str, int]] = (v for v in lst4)
    check(assert_type(df.set_index(iter1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.set_index(iter2), pd.DataFrame), pd.DataFrame)


def test_groupby_apply() -> None:
    # GH 167
    df = pd.DataFrame({"col1": [1, 2, 3], "col2": [4, 5, 6]})

    def sum_mean(x: pd.DataFrame) -> float:
        return x.sum().mean()

    check(assert_type(df.groupby("col1").apply(sum_mean), pd.Series), pd.Series)

    lfunc: Callable[[pd.DataFrame], float] = lambda x: x.sum().mean()
    check(assert_type(df.groupby("col1").apply(lfunc), pd.Series), pd.Series)

    def sum_to_list(x: pd.DataFrame) -> list:
        return x.sum().tolist()

    check(assert_type(df.groupby("col1").apply(sum_to_list), pd.Series), pd.Series)

    def sum_to_series(x: pd.DataFrame) -> pd.Series:
        return x.sum()

    check(
        assert_type(df.groupby("col1").apply(sum_to_series), pd.DataFrame), pd.DataFrame
    )

    def sample_to_df(x: pd.DataFrame) -> pd.DataFrame:
        return x.sample()

    check(
        assert_type(
            df.groupby("col1", group_keys=False).apply(sample_to_df), pd.DataFrame
        ),
        pd.DataFrame,
    )


def test_resample() -> None:
    # GH 181
    N = 10
    index = pd.date_range("1/1/2000", periods=N, freq="T")
    x = [x for x in range(N)]
    df = pd.DataFrame({"a": x, "b": x, "c": x}, index=index)
    check(assert_type(df.resample("2T").std(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").var(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").quantile(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").sum(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").prod(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").min(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").max(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").first(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").last(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").mean(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").sem(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").median(), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.resample("2T").ohlc(), pd.DataFrame), pd.DataFrame)


def test_loclist() -> None:
    # GH 189
    df = pd.DataFrame({1: [1, 2], None: 5}, columns=pd.Index([1, None], dtype=object))

    check(assert_type(df.loc[:, [None]], pd.DataFrame), pd.DataFrame)
    check(assert_type(df.loc[:, [1]], pd.DataFrame), pd.DataFrame)


def test_dict_items() -> None:
    # GH 180
    x = {"a": [1]}
    check(assert_type(pd.DataFrame(x.items()), pd.DataFrame), pd.DataFrame)


def test_generic() -> None:
    # GH 197
    T = TypeVar("T")

    class MyDataFrame(pd.DataFrame, Generic[T]):
        ...

    def func() -> MyDataFrame[int]:
        return MyDataFrame[int]({"foo": [1, 2, 3]})

    func()


def test_to_xarray():
    check(assert_type(DF.to_xarray(), xr.Dataset), xr.Dataset)


def test_to_records():
    check(assert_type(DF.to_records(False, "int8"), np.recarray), np.recarray)
    check(
        assert_type(DF.to_records(False, index_dtypes=np.int8), np.recarray),
        np.recarray,
    )
    check(
        assert_type(
            DF.to_records(False, {"col1": np.int8, "col2": np.int16}), np.recarray
        ),
        np.recarray,
    )


def test_to_dict():
    check(assert_type(DF.to_dict(), Dict[Hashable, Any]), dict)
    check(assert_type(DF.to_dict("split"), Dict[Hashable, Any]), dict)

    target: Mapping = defaultdict(list)
    check(assert_type(DF.to_dict(into=target), Mapping[Hashable, Any]), defaultdict)
    target = defaultdict(list)
    check(
        assert_type(DF.to_dict("tight", into=target), Mapping[Hashable, Any]),
        defaultdict,
    )
    target = defaultdict(list)
    check(assert_type(DF.to_dict("records"), List[Dict[Hashable, Any]]), list)
    check(
        assert_type(DF.to_dict("records", into=target), List[Mapping[Hashable, Any]]),
        list,
    )


def test_neg() -> None:
    # GH 253
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    check(assert_type(-df, pd.DataFrame), pd.DataFrame)


def test_pos() -> None:
    # GH 253
    df = pd.DataFrame(data={"col1": [1, 2], "col2": [3, 4]})
    check(assert_type(+df, pd.DataFrame), pd.DataFrame)


def test_getattr() -> None:
    # GH 261
    df = pd.DataFrame({"a": [1, 2]})
    check(assert_type(df.a, pd.Series), pd.Series)


def test_xs_key() -> None:
    # GH 214
    mi = pd.MultiIndex.from_product([[0, 1], [0, 1]], names=["foo", "bar"])
    df = pd.DataFrame({"x": [10, 20, 30, 40], "y": [50, 60, 70, 80]}, index=mi)
    check(
        assert_type(df.xs(0, level="foo"), Union[pd.DataFrame, pd.Series]), pd.DataFrame
    )


def test_loc_slice() -> None:
    # GH 277
    df1 = pd.DataFrame(
        {"x": [1, 2, 3, 4]},
        index=pd.MultiIndex.from_product([[1, 2], ["a", "b"]], names=["num", "let"]),
    )
    check(assert_type(df1.loc[1, :], Union[pd.Series, pd.DataFrame]), pd.DataFrame)


def test_where() -> None:
    df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})

    def cond1(x: int) -> bool:
        return x % 2 == 0

    check(assert_type(df.where(cond1), pd.DataFrame), pd.DataFrame)

    def cond2(x: pd.DataFrame) -> pd.DataFrame:
        return x > 1

    check(assert_type(df.where(cond2), pd.DataFrame), pd.DataFrame)

    cond3 = pd.DataFrame({"a": [True, True, False], "b": [False, False, False]})
    check(assert_type(df.where(cond3), pd.DataFrame), pd.DataFrame)


def test_setitem_loc() -> None:
    # GH 254
    df = pd.DataFrame.from_dict(
        {view: (True, True, True) for view in ["A", "B", "C"]}, orient="index"
    )
    df.loc[["A", "C"]] = False
    my_arr = ["A", "C"]
    df.loc[my_arr] = False


def test_replace_na() -> None:
    # GH 262
    frame = pd.DataFrame(["N/A", "foo", "bar"])
    check(assert_type(frame.replace("N/A", pd.NA), pd.DataFrame), pd.DataFrame)


def test_isetframe() -> None:
    frame = pd.DataFrame([[1, 2], [3, 4]], columns=["a", "b"])
    check(assert_type(frame.isetitem(0, 10), None), type(None))
    check(assert_type(frame.isetitem([0], [10, 12]), None), type(None))


def test_reset_index_150_changes() -> None:
    frame = pd.DataFrame({"a": [1, 2, 3, 4]}, index=[-10, -9, -8, -7])
    check(
        assert_type(
            frame.reset_index(allow_duplicates=True, names="idx"), pd.DataFrame
        ),
        pd.DataFrame,
    )
    check(
        assert_type(
            frame.reset_index(allow_duplicates=True, names=["idx"]), pd.DataFrame
        ),
        pd.DataFrame,
    )


def test_compare_150_changes() -> None:
    frame_a = pd.DataFrame({"a": [1, 2, 3, 4]}, index=[-10, -9, -8, -7])
    frame_b = pd.DataFrame({"a": [1, 2, 4, 3]}, index=[-10, -9, -8, -7])
    check(
        assert_type(
            frame_a.compare(frame_b, result_names=("one", "the_other")), pd.DataFrame
        ),
        pd.DataFrame,
    )


def test_quantile_150_changes() -> None:
    frame = pd.DataFrame(getSeriesData())
    check(assert_type(frame.quantile(0.5, method="single"), pd.Series), pd.Series)
    check(
        assert_type(
            frame.quantile([0.25, 0.5, 0.75], interpolation="nearest", method="table"),
            pd.DataFrame,
        ),
        pd.DataFrame,
    )


def test_resample_150_changes() -> None:
    idx = pd.date_range("2020-1-1", periods=700)
    frame = pd.DataFrame(np.random.standard_normal((700, 1)), index=idx, columns=["a"])
    resampler = frame.resample("M", group_keys=True)
    assert_type(resampler, "Resampler[pd.DataFrame]")

    def f(s: pd.DataFrame) -> pd.Series:
        return s.mean()

    check(assert_type(resampler.apply(f), Union[pd.Series, pd.DataFrame]), pd.DataFrame)


def test_df_accepting_dicts_iterator() -> None:
    # GH 392
    data = [{"a": 1, "b": 2}, {"a": 3, "b": 5}]
    check(assert_type(pd.DataFrame(iter(data)), pd.DataFrame), pd.DataFrame)


def test_series_added_in_astype() -> None:
    # GH410
    df = pd.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
    check(assert_type(df.astype(df.dtypes), pd.DataFrame), pd.DataFrame)


def test_series_groupby_and_value_counts() -> None:
    df = pd.DataFrame(
        {
            "Animal": ["Falcon", "Falcon", "Parrot", "Parrot"],
            "Max Speed": [380, 370, 24, 26],
        }
    )
    c1 = df.groupby("Animal")["Max Speed"].value_counts()
    c2 = df.groupby("Animal")["Max Speed"].value_counts(normalize=True)
    check(assert_type(c1, "pd.Series[int]"), pd.Series, np.int64)
    check(assert_type(c2, "pd.Series[float]"), pd.Series, float)


def test_axes_as_tuple() -> None:
    # GH 384
    index = (3, 5, 7)
    columns = ["a", "b", "c"]
    df = pd.DataFrame(data=1, index=index, columns=columns)
    check(assert_type(df, pd.DataFrame), pd.DataFrame)


def test_astype_dict() -> None:
    # GH 447
    df = pd.DataFrame({"a": [1, 2, 3], 43: [4, 5, 6]})
    columns_types = {"a": "int", 43: "float"}
    de = df.astype(columns_types)
    check(assert_type(de, pd.DataFrame), pd.DataFrame)
    check(assert_type(df.astype({"a": "int", 43: "float"}), pd.DataFrame), pd.DataFrame)


def test_setitem_none() -> None:
    df = pd.DataFrame(
        {"A": [1, 2, 3], "B": ["abc", "def", "ghi"]}, index=["x", "y", "z"]
    )
    df.loc["x", "B"] = None
    df.iloc[2, 0] = None
    sb = pd.Series([1, 2, 3], dtype=int)
    sb.loc["y"] = None
    sb.iloc[0] = None


def test_groupby_and_transform() -> None:
    df = pd.DataFrame(
        {
            "A": ["foo", "bar", "foo", "bar", "foo", "bar"],
            "B": ["one", "one", "two", "three", "two", "two"],
            "C": [1, 5, 5, 2, 5, 5],
            "D": [2.0, 5.0, 8.0, 1.0, 2.0, 9.0],
        }
    )
    ser = pd.Series(
        [390.0, 350.0, 30.0, 20.0],
        index=["Falcon", "Falcon", "Parrot", "Parrot"],
        name="Max Speed",
    )
    grouped = df.groupby("A")[["C", "D"]]
    grouped1 = ser.groupby(ser > 100)
    c1 = grouped.transform("sum")
    c2 = grouped.transform(lambda x: (x - x.mean()) / x.std())
    c3 = grouped1.transform("cumsum")
    c4 = grouped1.transform(lambda x: x.max() - x.min())
    check(assert_type(c1, pd.DataFrame), pd.DataFrame)
    check(assert_type(c2, pd.DataFrame), pd.DataFrame)
    check(assert_type(c3, pd.Series), pd.Series)
    check(assert_type(c4, pd.Series), pd.Series)


def test_getattr_and_dataframe_groupby() -> None:
    df = pd.DataFrame(
        data={"col1": [1, 1, 2], "col2": [3, 4, 5], "col3": [0, 1, 0], 0: [-1, -1, -1]}
    )
    check(assert_type(df.groupby("col1").col3.agg(min), pd.Series), pd.Series)
    check(
        assert_type(df.groupby("col1").col3.agg([min, max]), pd.DataFrame),
        pd.DataFrame,
    )


def test_getsetitem_multiindex() -> None:
    # GH 466
    rows = pd.Index(["project A", "project B", "project C"])
    years: tuple[str, ...] = ("Year 1", "Year 2", "Year 3")
    quarters: tuple[str, ...] = ("Q1", "Q2", "Q3", "Q4")
    index_tuples: list[tuple[str, ...]] = list(itertools.product(years, quarters))
    cols = pd.MultiIndex.from_tuples(index_tuples)
    budget = pd.DataFrame(index=rows, columns=cols)
    multi_index: tuple[str, str] = ("Year 1", "Q1")
    budget.loc["project A", multi_index] = 4700
    check(assert_type(budget.loc["project A", multi_index], Scalar), int)


def test_frame_dropna_subset() -> None:
    # GH 434
    data = {"col1": [1, 3, 4], "col2": [2, 3, 5], "col3": [2, 4, 4]}
    df = pd.DataFrame(data)
    check(
        assert_type(df.dropna(subset=df.columns.drop("col1")), pd.DataFrame),
        pd.DataFrame,
    )


def test_loc_callable() -> None:
    # GH 256
    df = pd.DataFrame({"x": [1, 2, 3], "y": [4, 5, 6]})

    def select1(df: pd.DataFrame) -> pd.Series:
        return df["x"] > 2.0

    check(assert_type(df.loc[select1], pd.DataFrame), pd.DataFrame)
    check(assert_type(df.loc[select1, :], pd.DataFrame), pd.DataFrame)

    def select2(df: pd.DataFrame) -> list[Hashable]:
        return [i for i in df.index if cast(int, i) % 2 == 1]

    check(assert_type(df.loc[select2, "x"], pd.Series), pd.Series)

    def select3(df: pd.DataFrame) -> int:
        return 1

    check(assert_type(df.loc[select3, "x"], Scalar), np.integer)


def test_npint_loc_indexer() -> None:
    # GH 508

    df = pd.DataFrame(dict(x=[1, 2, 3]), index=np.array([10, 20, 30], dtype="uint64"))

    def get_NDArray(df: pd.DataFrame, key: npt.NDArray[np.uint64]) -> pd.DataFrame:
        df2 = df.loc[key]
        return df2

    a: npt.NDArray[np.uint64] = np.array([10, 30], dtype="uint64")
    check(assert_type(get_NDArray(df, a), pd.DataFrame), pd.DataFrame)


def test_in_columns() -> None:
    # GH 532 (PR)
    df = pd.DataFrame(np.random.random((3, 4)), columns=["cat", "dog", "rat", "pig"])
    cols = [c for c in df.columns if "at" in c]
    check(assert_type(cols, list), list, str)
    check(assert_type(df.loc[:, cols], pd.DataFrame), pd.DataFrame)
    check(assert_type(df[cols], pd.DataFrame), pd.DataFrame)
    check(assert_type(df.groupby(by=cols).sum(), pd.DataFrame), pd.DataFrame)


def test_insert_newvalues() -> None:
    df = pd.DataFrame({"a": [1, 2]})
    ab = pd.DataFrame({"col1": [1, 2], "col2": [3, 4]})
    ef = pd.DataFrame({"z": [4, 5, 6]})
    assert assert_type(df.insert(loc=0, column="b", value=None), None) is None
    assert assert_type(ab.insert(loc=0, column="newcol", value=[99, 99]), None) is None
    assert assert_type(ef.insert(loc=0, column="g", value=4), None) is None


def test_astype() -> None:
    s = pd.DataFrame({"d": [1, 2]})
    ab = pd.DataFrame({"col1": [1, 2], "col2": ["a", "b"]})

    check(assert_type(s.astype(int), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype(pd.Int64Dtype()), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype(str), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype(bytes), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype(pd.Float64Dtype()), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype(complex), "pd.DataFrame"), pd.DataFrame)
    check(
        assert_type(ab.astype({"col1": "int32", "col2": str}), "pd.DataFrame"),
        pd.DataFrame,
    )
    check(assert_type(s.astype(pd.CategoricalDtype()), "pd.DataFrame"), pd.DataFrame)
    check(assert_type(s.astype("category"), "pd.DataFrame"), pd.DataFrame)  # GH 559

    population_dict = {
        "California": 38332521,
        "Texas": 26448193,
        "New York": 19651127,
        "Florida": 19552860,
        "Illinois": 12882135,
    }
    area_dict = {
        "California": 423967,
        "Texas": 695662,
        "New York": 141297,
        "Florida": 170312,
        "Illinois": 149995,
    }
    population = pd.Series(population_dict)
    area = pd.Series(area_dict)

    states = pd.DataFrame({"population": population, "area": area})
    check(assert_type(states.astype(object), pd.DataFrame), pd.DataFrame, object)


def test_xs_frame_new() -> None:
    d = {
        "num_legs": [4, 4, 2, 2],
        "num_wings": [0, 0, 2, 2],
        "class": ["mammal", "mammal", "mammal", "bird"],
        "animal": ["cat", "dog", "bat", "penguin"],
        "locomotion": ["walks", "walks", "flies", "walks"],
    }
    df = pd.DataFrame(data=d)
    df = df.set_index(["class", "animal", "locomotion"])
    s1 = df.xs("mammal", axis=0)
    s2 = df.xs("num_wings", axis=1)
    check(assert_type(s1, Union[pd.Series, pd.DataFrame]), pd.DataFrame)
    check(assert_type(s2, Union[pd.Series, pd.DataFrame]), pd.Series)


def test_align() -> None:
    df0 = pd.DataFrame(
        data=np.array(
            [
                ["A0", "A1", "A2", "A3"],
                ["B0", "B1", "B2", "B3"],
                ["C0", "C1", "C2", "C3"],
            ]
        ).T,
        index=[0, 1, 2, 3],
        columns=["A", "B", "C"],
    )

    s0 = pd.Series(data={1: "1", 3: "3", 5: "5"})
    aligned_df0, aligned_s0 = df0.align(s0, axis="index")
    check(assert_type(aligned_df0, pd.DataFrame), pd.DataFrame)
    check(assert_type(aligned_s0, pd.Series), pd.Series)

    s1 = pd.Series(data={"A": "A", "D": "D"})
    aligned_df0, aligned_s1 = df0.align(s1, axis="columns")
    check(assert_type(aligned_df0, pd.DataFrame), pd.DataFrame)
    check(assert_type(aligned_s1, pd.Series), pd.Series)

    df1 = pd.DataFrame(
        data=np.array(
            [
                ["A1", "A3", "A5"],
                ["D1", "D3", "D5"],
            ]
        ).T,
        index=[1, 3, 5],
        columns=["A", "D"],
    )
    aligned_df0, aligned_df1 = df0.align(df1)
    check(assert_type(aligned_df0, pd.DataFrame), pd.DataFrame)
    check(assert_type(aligned_df1, pd.DataFrame), pd.DataFrame)


def test_loc_returns_series() -> None:
    df1 = pd.DataFrame({"x": [1, 2, 3, 4]}, index=[10, 20, 30, 40])
    df2 = df1.loc[10, :]
    check(assert_type(df2, Union[pd.Series, pd.DataFrame]), pd.Series)


def test_to_dict_index() -> None:
    df = pd.DataFrame({"a": [1, 2], "b": [9, 10]})
    check(
        assert_type(
            df.to_dict(orient="records", index=True), List[Dict[Hashable, Any]]
        ),
        list,
    )
    check(assert_type(df.to_dict(orient="dict", index=True), Dict[Hashable, Any]), dict)
    check(
        assert_type(df.to_dict(orient="series", index=True), Dict[Hashable, Any]), dict
    )
    check(
        assert_type(df.to_dict(orient="index", index=True), Dict[Hashable, Any]), dict
    )
    check(
        assert_type(df.to_dict(orient="split", index=True), Dict[Hashable, Any]), dict
    )
    check(
        assert_type(df.to_dict(orient="tight", index=True), Dict[Hashable, Any]), dict
    )
    check(
        assert_type(df.to_dict(orient="tight", index=False), Dict[Hashable, Any]), dict
    )
    check(
        assert_type(df.to_dict(orient="split", index=False), Dict[Hashable, Any]), dict
    )
    if TYPE_CHECKING_INVALID_USAGE:
        check(assert_type(df.to_dict(orient="records", index=False), List[Dict[Hashable, Any]]), list)  # type: ignore[assert-type, call-overload] # pyright: ignore[reportGeneralTypeIssues]
        check(assert_type(df.to_dict(orient="dict", index=False), Dict[Hashable, Any]), dict)  # type: ignore[assert-type, call-overload] # pyright: ignore[reportGeneralTypeIssues]
        check(assert_type(df.to_dict(orient="series", index=False), Dict[Hashable, Any]), dict)  # type: ignore[assert-type, call-overload] # pyright: ignore[reportGeneralTypeIssues]
        check(assert_type(df.to_dict(orient="index", index=False), Dict[Hashable, Any]), dict)  # type: ignore[assert-type, call-overload] # pyright: ignore[reportGeneralTypeIssues]


def test_suffix_prefix_index() -> None:
    df = pd.DataFrame({"A": [1, 2, 3, 4], "B": [3, 4, 5, 6]})
    check(assert_type(df.add_suffix("_col", axis=1), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.add_suffix("_col", axis="index"), pd.DataFrame), pd.DataFrame)
    check(assert_type(df.add_prefix("_col", axis="index"), pd.DataFrame), pd.DataFrame)
    check(
        assert_type(df.add_prefix("_col", axis="columns"), pd.DataFrame), pd.DataFrame
    )


def test_convert_dtypes_dtype_backend() -> None:
    df = pd.DataFrame({"A": [1, 2, 3, 4], "B": [3, 4, 5, 6]})
    dfn = df.convert_dtypes(dtype_backend="numpy_nullable")
    check(assert_type(dfn, pd.DataFrame), pd.DataFrame)