Merge branch 'master' into release

Author: tmke8

.editorconfig

@@ -0,0 +1,14 @@
+# http://editorconfig.org
+
+root = true
+
+[*]
+indent_style = space
+indent_size = 4
+insert_final_newline = true
+trim_trailing_whitespace = true
+end_of_line = LF
+charset = utf-8
+
+[*.py]
+max_line_length = 100

.github/workflows/formatting.yml

@@ -25,8 +25,8 @@ jobs:
         python gen_pyi.py
     - name: Format with black
       run: |
-        python -m black --check -l 100 -t py36 -S .
-        python -m black --check -l 100 -t py36 -S --pyi --fast matplotlib-stubs/pyplot.pyi.in
+        python -m black --check -l 100 -t py36 .
+        python -m black --check -l 100 -t py36 --pyi --fast matplotlib-stubs/pyplot.pyi.in
     - name: Lint with flake8
       run: |
         flake8 *-stubs

.github/workflows/mypy_check.yml

@@ -19,7 +19,7 @@ jobs:
     - name: Install mypy
       run: |
         python -m pip install --upgrade pip
-        pip install mypy
+        pip install mypy==0.770
         pip install .
     - name: Check with mypy
       run: |

check_all.sh

@@ -1,3 +1,10 @@
 #!/bin/bash
 
-python gen_pyi.py && black -S -l 100 -t py36 . && flake8 *-stubs && python -m pytest -vv tests/ && mypy tests && echo "Check all complete!"
+echo "Begin check..." \
+&& black -l 100 -t py36 . \
+&& python gen_pyi.py \
+&& black -l 100 -t py36 --pyi --fast matplotlib-stubs/pyplot.pyi.in \
+&& flake8 *-stubs \
+&& python -m pytest -vv tests/ \
+&& mypy tests \
+&& echo "Check all complete!"

create_tag.py

@@ -10,39 +10,44 @@ def main() -> None:
     assert not repo.is_dirty(), "please commit or stash all other changes"
     assert repo.active_branch.name == "master", "only tag versions on master"
 
+    # merge master into release branch
+    git = repo.git
+    git.checkout("release")
+    git.merge("--no-ff", "master", "--no-edit")
+    print("updated release branch")
+
     # find out the version number
     tags_minor_versions = [tag.name.split(".")[-1] for tag in repo.tags]
-    minor_versions_as_ints = [int(mv) for mv in tags_minor_versions if mv.isdigit()]
-    latest_minor_version = max(minor_versions_as_ints)
+    latest_minor_version = max(int(mv) for mv in tags_minor_versions if mv.isdigit())
+    old_version = f"0.2.{latest_minor_version}"
     new_version = f"0.2.{latest_minor_version + 1}"
     print(f"new version will be: {new_version}")
 
     # change version in setup.py
     setup_path = Path("setup.py")
     with setup_path.open("r") as fhandle:
-        new_content = fhandle.read().replace("0.3.0.dev1", new_version)
+        file_content = fhandle.read()
+        count = 0
+        while old_version in file_content:
+            file_content = file_content.replace(old_version, new_version, 1)
+            count += 1
+        assert count == 1, "couldn't properly change the version in setup.py... aborting"
     with setup_path.open("w") as fhandle:
-        fhandle.write(new_content)
+        fhandle.write(file_content)
     print("modified setup.py")
 
     # commit change
-    git = repo.git
-    branch_name = f"releases/{new_version}"
-    print(f"commit to branch {branch_name}")
-    git.checkout("-b", branch_name)
     repo.index.add(["setup.py"])
-    repo.index.commit("Bump version")
+    repo.index.commit(f"Bump version to {new_version}")
 
-    # create tag and push
+    # # create tag and push
     tag_name = f"v{new_version}"
     repo.create_tag(tag_name)
-    print(f"git push origin {tag_name} {branch_name}")
-    git.push("origin", tag_name, branch_name)
+    print(f"git push origin release {tag_name}")
+    git.push("origin", "release", tag_name)
 
     # clean up
     git.checkout("master")
-    git.branch("-D", branch_name)
-    print(f"Deleted branch {branch_name}")
 
 
 if __name__ == "__main__":

pandas-stubs/__init__.pyi

@@ -52,7 +52,7 @@ def read_sql(
 def read_feather(p: Union[Path, IO]) -> DataFrame: ...
 def to_numeric(
     arg: Union[int, float, List, Tuple, _np.ndarray, Series],
-    errors: Literal['ignore', 'raise', 'coerce'] = ...,
-    downcast: Literal['integer', 'signed', 'unsigned', 'float'] = ...,
+    errors: Literal["ignore", "raise", "coerce"] = ...,
+    downcast: Literal["integer", "signed", "unsigned", "float"] = ...,
 ) -> Union[Series, _np.ndarray]: ...
 def unique(values: Series) -> _np.ndarray: ...

pandas-stubs/core/frame.pyi

@@ -37,10 +37,12 @@ _DType = TypeVar("_DType", bound=_np.dtype)
 
 _ColSubsetType = Union[Series, DataFrame, List[_str], _str, _np.ndarray[_np.str_]]
 
+_FunctionLike = Union[_str, Callable]
+
 class DataFrame:
     def __init__(
         self,
-        data: Optional[Union[_ListLike, DataFrame, Dict[_str, _np.ndarray]]] = ...,
+        data: Optional[Union[_ListLike, DataFrame, Dict[_str, _ListLike]]] = ...,
         columns: Optional[_ListLike] = ...,
         index: Optional[_ListLike] = ...,
         dtype: Optional[Type[_np.dtype]] = ...,
@@ -55,6 +57,7 @@ class DataFrame:
     def __getitem__(
         self, idx: Union[Series, DataFrame, List[_str], Index[_str], _np.ndarray[_np.str_]]
     ) -> DataFrame: ...
+    def __getattr__(self, name: _str) -> Series: ...
     def __iter__(self) -> Iterator: ...
     def __len__(self) -> int: ...
     def __setitem__(self, key: Any, value: Any) -> None: ...
@@ -164,9 +167,33 @@ class DataFrame:
         downcast: Dict = ...,
     ) -> None: ...
     @overload
-    def groupby(self, by: List[_str]) -> DataFrameGroupBy: ...
+    def groupby(
+        self,
+        by: Union[
+            _str,
+            Tuple[_str, ...],
+            List[_str],
+            List[Tuple[_str, _str]],
+            List[Tuple[_str, _str, _str]],
+        ],
+        level: Union[int, _str] = ...,
+        as_index: bool = ...,
+        sort: bool = ...,
+        group_keys: bool = ...,
+        squeeze: bool = ...,
+        observed: bool = ...,
+    ) -> DataFrameGroupBy: ...
     @overload
-    def groupby(self, by: _str) -> SeriesGroupBy: ...
+    def groupby(
+        self,
+        by: Union[Series[_str], Dict[_str, _str], Callable],
+        axis: _AxisType = ...,
+        level: Union[int, _str] = ...,
+        sort: bool = ...,
+        group_keys: bool = ...,
+        squeeze: bool = ...,
+        observed: bool = ...,
+    ) -> DataFrameGroupBy: ...
     def head(self, n: int) -> DataFrame: ...
     def idxmax(self, axis: _AxisType) -> Series: ...
     def insert(

pandas-stubs/core/groupby/generic.pyi

@@ -1,21 +1,50 @@
-from typing import overload, List, Iterator
+from typing import overload, Optional, Union, List, Dict, Iterator
+import numpy as _np
 
 from pandas.core.frame import DataFrame
 from pandas.core.series import Series
+from ..frame import _FunctionLike
+
+_str = str  # needed because Series has a property called "str"...
 
 class GroupBy: ...
 
 class DataFrameGroupBy(GroupBy):
     @overload
-    def __getitem__(self, item: str) -> Series: ...
+    def __getitem__(self, item: _str) -> SeriesGroupBy: ...
     @overload
-    def __getitem__(self, item: List[str]) -> DataFrame: ...
+    def __getitem__(self, item: List[_str]) -> DataFrameGroupBy: ...
+    def __getattr__(self, name: _str) -> SeriesGroupBy: ...
     def __iter__(self) -> Iterator: ...
+    def aggregate(
+        self, func: Union[_FunctionLike, List[_FunctionLike], Dict[_str, _FunctionLike]],
+    ) -> DataFrame: ...
+    agg = aggregate
+    def count(self) -> DataFrame: ...
+    def head(self, n: int = ...) -> DataFrame: ...
+    def max(self) -> DataFrame: ...
+    def mean(self) -> DataFrame: ...
+    def median(self) -> DataFrame: ...
+    def min(self) -> DataFrame: ...
+    def nunique(self, dropna: bool = ...) -> DataFrame: ...
+    def quantile(self, q: float = ..., interpolation: str = ...) -> DataFrame: ...
+    def std(self, ddof: int = ...) -> DataFrame: ...
+    def sum(self) -> DataFrame: ...
+    def tail(self, n: int = ...) -> DataFrame: ...
+    def var(self, ddof: int = ...) -> DataFrame: ...
 
 class SeriesGroupBy(GroupBy):
-    def __getitem__(self, item: str) -> Series: ...
+    def __getitem__(self, item: _str) -> Series: ...
+    def count(self) -> Series: ...
+    def head(self, n: int = ...) -> Series: ...
     def max(self) -> Series: ...
     def mean(self) -> Series: ...
     def median(self) -> Series: ...
     def min(self) -> Series: ...
+    def nunique(self, dropna: bool = ...) -> Series: ...
+    def quantile(self, q: float = ..., interpolation: str = ...) -> Series: ...
     def std(self, ddof: int = ...) -> Series: ...
+    def sum(self) -> Series: ...
+    def tail(self, n: int = ...) -> Series: ...
+    def unique(self) -> Series[_np.ndarray]: ...
+    def var(self, ddof: int = ...) -> Series: ...

pandas-stubs/core/series.pyi

@@ -25,7 +25,7 @@ from .strings import StringMethods
 
 _str = str  # needed because Series has a property called "str"...
 
-_DType = TypeVar("_DType", _str, bool, int, float, object)
+_DType = TypeVar("_DType", _str, bool, int, float, object, _np.ndarray, List)
 _ListLike = Union[_np.ndarray, List[_DType], Dict[_str, _np.ndarray]]
 # dtypes for numpy
 _DTypeNp = TypeVar(

tests/pandas_test.py

@@ -21,6 +21,25 @@ def assert_type(v: T, t: Type[T]) -> None:
     [[1.0, 2.0], [4.0, 5.0], [7.0, 8.0]], columns=["max_speed", "shield"],
 )
 s: pd.Series[float] = df["shield"].copy()
+iris = pd.DataFrame(
+    {
+        "sepal_length": [5.1, 4.9, 4.7, 7.0, 6.4, 6.9, 6.3, 5.8, 7.1],
+        "sepal_width": [3.5, 3.0, 3.2, 3.2, 3.2, 3.1, 3.3, 2.7, 3.0],
+        "petal_length": [1.4, 1.4, 1.3, 4.7, 4.5, 4.9, 6.0, 5.1, 5.9],
+        "petal_width": [0.2, 0.2, 0.2, 1.4, 1.5, 1.5, 2.5, 1.9, 2.1],
+        "species": [
+            "setosa",
+            "setosa",
+            "setosa",
+            "versicolor",
+            "versicolor",
+            "versicolor",
+            "virginica",
+            "virginica",
+            "virginica",
+        ],
+    }
+)
 
 
 def test_getitem() -> None:
@@ -85,3 +104,17 @@ def test_series_iloc() -> None:
 def test_multiindex() -> None:
     tuples = [("bar", "one"), ("bar", "two"), ("baz", "one")]
     index: pd.MultiIndex = pd.MultiIndex.from_tuples(tuples, names=["first", "second"])
+
+
+def test_iris() -> None:
+    sepal_length = iris.groupby("species")["sepal_length"]
+    arr: np.ndarray = sepal_length.unique()[0]
+    assert isinstance(arr, np.ndarray)
+    d: pd.DataFrame = iris.groupby("species", as_index=False).max()
+    assert_type(d, pd.DataFrame)
+    e: pd.DataFrame = iris.groupby(
+        {"sepal_length": "length", "petal_length": "length"}, axis=1
+    ).agg(lambda x: x.mean(axis="columns"))
+    assert_type(e, pd.DataFrame)
+    f: pd.Series = e["length"]
+    assert_type(f, pd.Series)