Merge branch 'main' into add-where-expression

.github/release-drafter.yml

@@ -9,31 +9,31 @@ autolabeler:
   - label: breaking
     title:
       # Example: feat!: ...
-      - '/^(build|chore|ci|depr|docs|feat|fix|perf|refactor|release|test)(\(.*\))?\!\: /'
+      - '/^([B|b]uild|[C|c]hore|[CI|ci]|[D|d]epr|[D|d]oc|DOC|[F|f]eat|[F|f]ix|[P|p]erf|[R|r]efactor|[R|r]elease|[T|t]est)\! /'
   - label: build
     title:
-      - '/^build/'
+      - '/^[B|b]uild/'
   - label: internal
     title:
-      - '/^(chore|ci|refactor|test|template|bench)/'
+      - '/^[C|c]hore|[CI|ci]|[R|r]efactor|[T|t]est|[T|t]emplate|[B|b]ench/'
   - label: deprecation
     title:
-      - '/^depr/'
+      - '/^[D|d]epr/'
   - label: documentation
     title:
-      - '/(.*doc|.*docstring)/'
+      - '/^[D|d]oc|DOC/'
   - label: enhancement
     title:
-      - '/^(.*feat|.*enh)/'
+      - '/^.*feat|.*enh|Feat|ENH|Enh/'
   - label: fix
     title:
-      - '/^fix/'
+      - '/^[F|f]ix/'
   - label: performance
     title:
-      - '/^perf/'
+      - '/^[P|p]erf/'
   - label: release
     title:
-      - '/^release/'
+      - '/^[R|r]elease/'
 
 version-resolver:
   major:

.github/workflows/extremes.yml

@@ -99,8 +99,8 @@ jobs:
         run: uv pip install --upgrade tox virtualenv setuptools pip -r requirements-dev.txt --system
       - name: uninstall pyarrow
         run: uv pip uninstall pyarrow --system
-      - name: install pyarrow nightly
-        run: uv pip install --pre --extra-index-url https://pypi.fury.io/arrow-nightlies/ pyarrow --system
+      # - name: install pyarrow nightly
+      #   run: uv pip install --extra-index-url https://pypi.fury.io/arrow-nightlies/ --pre pyarrow --system
       - name: uninstall pandas
         run: uv pip uninstall pandas --system
       - name: install-pandas-nightly

docs/installation.md

@@ -11,6 +11,6 @@ Then, if you start the Python REPL and see the following:
 ```python
 >>> import narwhals
 >>> narwhals.__version__
-'1.4.2'
+'1.5.0'
 ```
 then installation worked correctly!

narwhals/__init__.py

@@ -50,7 +50,7 @@
 from narwhals.utils import maybe_convert_dtypes
 from narwhals.utils import maybe_set_index
 
-__version__ = "1.4.2"
+__version__ = "1.5.0"
 
 __all__ = [
     "selectors",

narwhals/_dask/expr.py

@@ -475,7 +475,9 @@ def fill_null(self, value: Any) -> DaskExpr:
         )
 
     def clip(
-        self: Self, lower_bound: Any | None = None, upper_bound: Any | None = None
+        self: Self,
+        lower_bound: Any | None = None,
+        upper_bound: Any | None = None,
     ) -> Self:
         return self._from_call(
             lambda _input, _lower, _upper: _input.clip(lower=_lower, upper=_upper),
@@ -545,6 +547,40 @@ def func(_input: Any) -> Any:
             returns_scalar=False,
         )
 
+    def is_duplicated(self: Self) -> Self:
+        def func(_input: Any) -> Any:
+            _name = _input.name
+            return (
+                _input.to_frame().groupby(_name).transform("size", meta=(_name, int)) > 1
+            )
+
+        return self._from_call(
+            func,
+            "is_duplicated",
+            returns_scalar=False,
+        )
+
+    def is_unique(self: Self) -> Self:
+        def func(_input: Any) -> Any:
+            _name = _input.name
+            return (
+                _input.to_frame().groupby(_name).transform("size", meta=(_name, int)) == 1
+            )
+
+        return self._from_call(
+            func,
+            "is_unique",
+            returns_scalar=False,
+        )
+
+    def is_in(self: Self, other: Any) -> Self:
+        return self._from_call(
+            lambda _input, other: _input.isin(other),
+            "is_in",
+            other,
+            returns_scalar=False,
+        )
+
     def null_count(self: Self) -> Self:
         return self._from_call(
             lambda _input: _input.isna().sum(),
@@ -764,6 +800,49 @@ def ordinal_day(self) -> DaskExpr:
             returns_scalar=False,
         )
 
+    def to_string(self, format: str) -> DaskExpr:  # noqa: A002
+        return self._expr._from_call(
+            lambda _input, _format: _input.dt.strftime(_format),
+            "strftime",
+            format.replace("%.f", ".%f"),
+            returns_scalar=False,
+        )
+
+    def total_minutes(self) -> DaskExpr:
+        return self._expr._from_call(
+            lambda _input: _input.dt.total_seconds() // 60,
+            "total_minutes",
+            returns_scalar=False,
+        )
+
+    def total_seconds(self) -> DaskExpr:
+        return self._expr._from_call(
+            lambda _input: _input.dt.total_seconds() // 1,
+            "total_seconds",
+            returns_scalar=False,
+        )
+
+    def total_milliseconds(self) -> DaskExpr:
+        return self._expr._from_call(
+            lambda _input: _input.dt.total_seconds() * 1000 // 1,
+            "total_milliseconds",
+            returns_scalar=False,
+        )
+
+    def total_microseconds(self) -> DaskExpr:
+        return self._expr._from_call(
+            lambda _input: _input.dt.total_seconds() * 1_000_000 // 1,
+            "total_microseconds",
+            returns_scalar=False,
+        )
+
+    def total_nanoseconds(self) -> DaskExpr:
+        return self._expr._from_call(
+            lambda _input: _input.dt.total_seconds() * 1_000_000_000 // 1,
+            "total_nanoseconds",
+            returns_scalar=False,
+        )
+
 
 class DaskExprNameNamespace:
     def __init__(self: Self, expr: DaskExpr) -> None:

pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "narwhals"
-version = "1.4.2"
+version = "1.5.0"
 authors = [
   { name="Marco Gorelli", email="[email protected]" },
 ]
@@ -26,6 +26,8 @@ exclude = [
 ]
 
 [project.optional-dependencies]
+cudf = ["cudf>=23.08.00"]
+modin = ["modin"]
 pandas = ["pandas>=0.25.3"]
 polars = ["polars>=0.20.3"]
 pyarrow = ['pyarrow>=11.0.0']
@@ -94,6 +96,7 @@ filterwarnings = [
   'ignore:.*Do not use the `random` module inside strategies',
   'ignore:.*You are using pyarrow version',
   'ignore:.*but when imported by',
+  'ignore:Distributing .*This may take some time',
 ]
 xfail_strict = true
 markers = ["slow: marks tests as slow (deselect with '-m \"not slow\"')"]

tests/expr_and_series/arithmetic_test.py

@@ -143,10 +143,10 @@ def test_truediv_same_dims(constructor_eager: Any, request: Any) -> None:
         request.applymarker(pytest.mark.xfail)
     s_left = nw.from_native(constructor_eager({"a": [1, 2, 3]}), eager_only=True)["a"]
     s_right = nw.from_native(constructor_eager({"a": [2, 2, 1]}), eager_only=True)["a"]
-    result = (s_left / s_right).to_list()
-    assert result == [0.5, 1.0, 3.0]
-    result = (s_left.__rtruediv__(s_right)).to_list()
-    assert result == [2, 1, 1 / 3]
+    result = s_left / s_right
+    compare_dicts({"a": result}, {"a": [0.5, 1.0, 3.0]})
+    result = s_left.__rtruediv__(s_right)
+    compare_dicts({"a": result}, {"a": [2, 1, 1 / 3]})
 
 
 @pytest.mark.slow()

tests/expr_and_series/cat/get_categories_test.py

@@ -25,8 +25,8 @@ def test_get_categories(request: Any, constructor_eager: Any) -> None:
     result_expr = df.select(nw.col("a").cat.get_categories())
     compare_dicts(result_expr, expected)
 
-    result_series = df["a"].cat.get_categories().to_list()
-    assert set(result_series) == set(expected["a"])
+    result_series = df["a"].cat.get_categories()
+    compare_dicts({"a": result_series}, expected)
 
 
 def test_get_categories_pyarrow() -> None:
@@ -41,5 +41,5 @@ def test_get_categories_pyarrow() -> None:
     result_expr = df.select(nw.col("a").cat.get_categories())
     compare_dicts(result_expr, expected)
 
-    result_series = df["a"].cat.get_categories().to_list()
-    assert result_series == expected["a"]
+    result_series = df["a"].cat.get_categories()
+    compare_dicts({"a": result_series}, expected)

tests/expr_and_series/dt/datetime_duration_test.py

@@ -44,9 +44,7 @@ def test_duration_attributes(
     expected_b: list[int],
     expected_c: list[int],
 ) -> None:
-    if "dask_lazy" in str(constructor) or (
-        parse_version(pd.__version__) < (2, 2) and "pandas_pyarrow" in str(constructor)
-    ):
+    if parse_version(pd.__version__) < (2, 2) and "pandas_pyarrow" in str(constructor):
         request.applymarker(pytest.mark.xfail)
 
     df = nw.from_native(constructor(data))