dataframely 1.0.0__tar.gz → 1.2.0__tar.gz

{dataframely-1.0.0 → dataframely-1.2.0}/.github/workflows/build.yml

@@ -15,7 +15,7 @@ jobs:
         with:
           fetch-depth: 0
       - name: Set up pixi
-        uses: prefix-dev/setup-pixi@0f64e482e3d251f735019b1bc7fb0413ead75b2c # v0.8.2
+        uses: prefix-dev/setup-pixi@19eac09b398e3d0c747adc7921926a6d802df4da # v0.8.8
         with:
           environments: build
       - name: Set version
@@ -23,7 +23,7 @@ jobs:
       - name: Build project
         run: pixi run -e build build-sdist
       - name: Upload package
-        uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
+        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
           name: sdist
           path: dist/*
@@ -50,7 +50,7 @@ jobs:
         with:
           fetch-depth: 0
       - name: Set up pixi
-        uses: prefix-dev/setup-pixi@0f64e482e3d251f735019b1bc7fb0413ead75b2c # v0.8.2
+        uses: prefix-dev/setup-pixi@19eac09b398e3d0c747adc7921926a6d802df4da # v0.8.8
         with:
           environments: build
       - name: Set version
@@ -64,7 +64,7 @@ jobs:
       - name: Check package
         run: pixi run -e build check-wheel
       - name: Upload package
-        uses: actions/upload-artifact@65c4c4a1ddee5b72f698fdd19549f0f0fb45cf08 # v4.6.0
+        uses: actions/upload-artifact@ea165f8d65b6e75b540449e92b4886f43607fa02 # v4.6.2
         with:
           name: wheel-${{ matrix.target-platform }}
           path: dist/*
@@ -78,7 +78,7 @@ jobs:
       id-token: write
     environment: pypi
     steps:
-      - uses: actions/download-artifact@fa0a91b85d4f404e444e00e005971372dc801d16 # v4.1.8
+      - uses: actions/download-artifact@95815c38cf2ff2164869cbab79da8d1f422bc89e # v4.2.1
         with:
           path: dist
           merge-multiple: true

{dataframely-1.0.0 → dataframely-1.2.0}/.github/workflows/chore.yml

@@ -1,6 +1,6 @@
 name: Chore
 on:
-  pull_request:
+  pull_request_target:
     branches: [main]
     types: [opened, reopened, edited, synchronize]
   push:
@@ -13,7 +13,7 @@ concurrency:
 jobs:
   check-pr-title:
     name: Check PR Title
-    if: github.event_name == 'pull_request'
+    if: github.event_name == 'pull_request_target'
     runs-on: ubuntu-latest
     permissions:
       contents: read
@@ -51,16 +51,16 @@ jobs:
           delete: true
 
   release-drafter:
-    name: ${{ github.event_name == 'pull_request' && 'Assign Labels' || 'Draft Release' }}
+    name: ${{ github.event_name == 'pull_request_target' && 'Assign Labels' || 'Draft Release' }}
     runs-on: ubuntu-latest
     permissions:
       contents: write
       pull-requests: write
     steps:
-      - name: ${{ github.event_name == 'pull_request' && 'Assign labels' || 'Update release draft' }}
+      - name: ${{ github.event_name == 'pull_request_target' && 'Assign labels' || 'Update release draft' }}
         uses: release-drafter/release-drafter@v6
         with:
-          disable-releaser: ${{ github.event_name == 'pull_request' }}
+          disable-releaser: ${{ github.event_name == 'pull_request_target' }}
           disable-autolabeler: ${{ github.event_name == 'push' }}
         env:
           GITHUB_TOKEN: ${{ github.token }}

{dataframely-1.0.0 → dataframely-1.2.0}/.github/workflows/ci.yml

@@ -21,7 +21,7 @@ jobs:
           # needed for 'pre-commit-mirrors-insert-license'
           fetch-depth: 0
       - name: Set up pixi
-        uses: prefix-dev/setup-pixi@0f64e482e3d251f735019b1bc7fb0413ead75b2c # v0.8.2
+        uses: prefix-dev/setup-pixi@19eac09b398e3d0c747adc7921926a6d802df4da # v0.8.8
         with:
           environments: default lint
       - name: Install repository
@@ -42,7 +42,7 @@ jobs:
       - name: Checkout branch
         uses: actions/checkout@v4
       - name: Set up pixi
-        uses: prefix-dev/setup-pixi@8eaba7c61d661f73d558b0b477156b7b62667fa4
+        uses: prefix-dev/setup-pixi@19eac09b398e3d0c747adc7921926a6d802df4da # v0.8.8
         with:
           environments: ${{ matrix.environment }}
       - name: Install repository
@@ -53,3 +53,4 @@ jobs:
         uses: codecov/codecov-action@v5
         with:
           files: ./coverage.xml
+          token: ${{ secrets.CODECOV_TOKEN }}

{dataframely-1.0.0 → dataframely-1.2.0}/.github/workflows/update-lockfiles.yml

@@ -15,13 +15,13 @@ jobs:
     steps:
       - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683 # v4.2.2
       - name: Set up pixi
-        uses: prefix-dev/setup-pixi@0f64e482e3d251f735019b1bc7fb0413ead75b2c # v0.8.2
+        uses: prefix-dev/setup-pixi@19eac09b398e3d0c747adc7921926a6d802df4da # v0.8.8
         with:
           run-install: false
       - name: Update lockfiles
         run: pixi update --json --no-install | pixi exec pixi-diff-to-markdown >> diff.md
       - name: Create pull request
-        uses: peter-evans/create-pull-request@67ccf781d68cd99b580ae25a5c18a1cc84ffff1f # v7.0.6
+        uses: peter-evans/create-pull-request@271a8d0340265f705b14b6d32b9829c1cb33d45e # v7.0.8
         with:
           token: ${{ github.token }}
           commit-message: Update pixi lockfile

{dataframely-1.0.0 → dataframely-1.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dataframely
-Version: 1.0.0
+Version: 1.2.0
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
@@ -12,7 +12,8 @@ Summary: A declarative, polars-native data frame validation library
 Author-email: Andreas Albert <andreas.albert@quantco.com>, Daniel Elsner <daniel.elsner@quantco.com>, Oliver Borchert <oliver.borchert@quantco.com>
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
-Project-URL: repository, https://github.com/quantco/dataframely
+Project-URL: Repository, https://github.com/quantco/dataframely
+Project-URL: Documentation, https://dataframely.readthedocs.io/
 
 <!-- LOGO -->
 <br />

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/__init__.py

@@ -34,6 +34,7 @@ from .columns import (
     Int64,
     Integer,
     List,
+    Object,
     String,
     Struct,
     Time,
@@ -89,4 +90,5 @@ __all__ = [
     "String",
     "Struct",
     "List",
+    "Object",
 ]

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/_base_collection.py

@@ -7,7 +7,7 @@ import typing
 from abc import ABCMeta
 from collections.abc import Iterable
 from dataclasses import dataclass, field
-from typing import Annotated, Any, Self, get_args, get_origin
+from typing import Annotated, Any, Self, cast, get_args, get_origin
 
 import polars as pl
 
@@ -48,6 +48,12 @@ class CollectionMember:
 
     #: Whether the member should be ignored in the filter method.
     ignored_in_filters: bool = False
+    #: Whether the member's non-primary key columns should be inlined for sampling.
+    #: This means that value overrides are supplied on the top-level rather than in
+    #: a subkey with the member's name. Only valid if the member's primary key matches
+    #: the collection's common primary key. Two members that share common column names
+    #: may not both be inlined for sampling.
+    inline_for_sampling: bool = False
 
 
 # --------------------------------------- UTILS -------------------------------------- #
@@ -79,7 +85,7 @@ class Metadata:
     members: dict[str, MemberInfo] = field(default_factory=dict)
     filters: dict[str, Filter] = field(default_factory=dict)
 
-    def update(self, other: Self):
+    def update(self, other: Self) -> None:
         self.members.update(other.members)
         self.filters.update(other.filters)
 
@@ -92,7 +98,7 @@ class CollectionMeta(ABCMeta):
         namespace: dict[str, Any],
         *args: Any,
         **kwargs: Any,
-    ):
+    ) -> CollectionMeta:
         result = Metadata()
         for base in bases:
             result.update(mcs._get_metadata_recursively(base))
@@ -136,6 +142,30 @@ class CollectionMeta(ABCMeta):
                     f"{len(intersection)} such filters: {sorted(intersection)}."
                 )
 
+        # 3) Check that inlining for sampling is configured correctly.
+        if len(non_ignored_member_schemas) > 0:
+            common_primary_keys = _common_primary_keys(non_ignored_member_schemas)
+            inlined_columns: set[str] = set()
+            for member, info in result.members.items():
+                if info.inline_for_sampling:
+                    if set(info.schema.primary_keys()) != common_primary_keys:
+                        raise ImplementationError(
+                            f"Member '{member}' is inlined for sampling but its primary "
+                            "key is a superset of the common primary key. Such a member "
+                            "must not be inlined to be able to provide multiple values "
+                            "for a single combination of the common primary key."
+                        )
+                    non_primary_key_columns = (
+                        set(info.schema.column_names()) - common_primary_keys
+                    )
+                    if len(inlined_columns & non_primary_key_columns):
+                        raise ImplementationError(
+                            f"At least one column name of member '{member}' clashes "
+                            "with a column name of another member that is inlined for "
+                            "sampling."
+                        )
+                    inlined_columns.update(non_primary_key_columns)
+
         return super().__new__(mcs, name, bases, namespace, *args, **kwargs)
 
     @staticmethod
@@ -153,58 +183,9 @@ class CollectionMeta(ABCMeta):
         # Get all members via the annotations
         if "__annotations__" in source:
             for attr, kls in source["__annotations__"].items():
-                origin = get_origin(kls)
-
-                # optional annotation
-                collection_member = CollectionMember()
-
-                if origin is Annotated:
-                    annotation_args = get_args(kls)
-                    origin_arg0 = get_origin(annotation_args[0])
-                    if not origin_arg0 or not issubclass(origin_arg0, TypedLazyFrame):
-                        raise AnnotationImplementationError(attr, kls)
-                    if len(annotation_args) > 2:
-                        raise AnnotationImplementationError(attr, kls)
-                    if not isinstance(annotation_args[1], CollectionMember):
-                        raise AnnotationImplementationError(attr, kls)
-
-                    # Continue with wrapped FrameType
-                    collection_member = annotation_args[1]
-                    kls = annotation_args[0]
-                    origin = origin_arg0
-
-                if origin is None:
-                    # `None` annotation is not allowed
-                    raise AnnotationImplementationError(attr, kls)
-                elif origin == typing.Union:
-                    # Happy path: optional member
-                    union_args = get_args(kls)
-                    if len(union_args) != 2:
-                        raise AnnotationImplementationError(attr, kls)
-                    if not any(get_origin(arg) is None for arg in union_args):
-                        raise AnnotationImplementationError(attr, kls)
-
-                    [not_none_arg] = [
-                        arg for arg in union_args if get_origin(arg) is not None
-                    ]
-                    if not issubclass(get_origin(not_none_arg), TypedLazyFrame):
-                        raise AnnotationImplementationError(attr, kls)
-
-                    result.members[attr] = MemberInfo(
-                        schema=get_args(not_none_arg)[0],
-                        is_optional=True,
-                        ignored_in_filters=collection_member.ignored_in_filters,
-                    )
-                elif issubclass(origin, TypedLazyFrame):
-                    # Happy path: required member
-                    result.members[attr] = MemberInfo(
-                        schema=get_args(kls)[0],
-                        is_optional=False,
-                        ignored_in_filters=collection_member.ignored_in_filters,
-                    )
-                else:
-                    # Some other unknown annotation
-                    raise AnnotationImplementationError(attr, kls)
+                result.members[attr] = CollectionMeta._derive_member_info(
+                    attr, kls, CollectionMember()
+                )
 
         # Get all filters by traversing the source
         for attr, value in {
@@ -215,6 +196,55 @@ class CollectionMeta(ABCMeta):
 
         return result
 
+    @staticmethod
+    def _derive_member_info(
+        attr: str, type_annotation: Any, collection_member: CollectionMember
+    ) -> MemberInfo:
+        origin = get_origin(type_annotation)
+
+        if origin is None:
+            # `None` annotation is not allowed
+            raise AnnotationImplementationError(attr, type_annotation)
+        elif origin == Annotated:
+            # Maybe happy path: annotated member, dispatch recursively
+            annotation_args = cast(list[Any], get_args(type_annotation))
+            if len(annotation_args) > 2:
+                raise AnnotationImplementationError(attr, type_annotation)
+            if not isinstance(annotation_args[1], CollectionMember):
+                raise AnnotationImplementationError(attr, type_annotation)
+            return CollectionMeta._derive_member_info(
+                attr, annotation_args[0], annotation_args[1]
+            )
+        elif origin == typing.Union:
+            # Happy path: optional member
+            union_args = get_args(type_annotation)
+            if len(union_args) != 2:
+                raise AnnotationImplementationError(attr, type_annotation)
+            if not any(get_origin(arg) is None for arg in union_args):
+                raise AnnotationImplementationError(attr, type_annotation)
+
+            [not_none_arg] = [arg for arg in union_args if get_origin(arg) is not None]
+            if not issubclass(get_origin(not_none_arg), TypedLazyFrame):
+                raise AnnotationImplementationError(attr, type_annotation)
+
+            return MemberInfo(
+                schema=get_args(not_none_arg)[0],
+                is_optional=True,
+                ignored_in_filters=collection_member.ignored_in_filters,
+                inline_for_sampling=collection_member.inline_for_sampling,
+            )
+        elif issubclass(origin, TypedLazyFrame):
+            # Happy path: required member
+            return MemberInfo(
+                schema=get_args(type_annotation)[0],
+                is_optional=False,
+                ignored_in_filters=collection_member.ignored_in_filters,
+                inline_for_sampling=collection_member.inline_for_sampling,
+            )
+        else:
+            # Some other unknown annotation
+            raise AnnotationImplementationError(attr, type_annotation)
+
 
 class BaseCollection(metaclass=CollectionMeta):
     """Internal utility abstraction to reference collections without introducing

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/_base_schema.py

@@ -58,7 +58,7 @@ class Metadata:
     columns: dict[str, Column] = field(default_factory=dict)
     rules: dict[str, Rule] = field(default_factory=dict)
 
-    def update(self, other: Self):
+    def update(self, other: Self) -> None:
         self.columns.update(other.columns)
         self.rules.update(other.rules)
 
@@ -71,7 +71,7 @@ class SchemaMeta(ABCMeta):
         namespace: dict[str, Any],
         *args: Any,
         **kwargs: Any,
-    ):
+    ) -> SchemaMeta:
         result = Metadata()
         for base in bases:
             result.update(mcs._get_metadata_recursively(base))

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/_compat.py

@@ -6,7 +6,7 @@ from typing import Any
 
 
 class _DummyModule:  # pragma: no cover
-    def __init__(self, module: str):
+    def __init__(self, module: str) -> None:
         self.module = module
 
     def __getattr__(self, name: str) -> Any:

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/_filter.py

@@ -12,7 +12,7 @@ C = TypeVar("C")
 class Filter(Generic[C]):
     """Internal class representing logic for filtering members of a collection."""
 
-    def __init__(self, logic: Callable[[C], pl.LazyFrame]):
+    def __init__(self, logic: Callable[[C], pl.LazyFrame]) -> None:
         self.logic = logic
 
 

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/_rule.py

@@ -12,14 +12,14 @@ ValidationFunction = Callable[[], pl.Expr]
 class Rule:
     """Internal class representing validation rules."""
 
-    def __init__(self, expr: pl.Expr):
+    def __init__(self, expr: pl.Expr) -> None:
         self.expr = expr
 
 
 class GroupRule(Rule):
     """Rule that is evaluated on a group of columns."""
 
-    def __init__(self, expr: pl.Expr, group_columns: list[str]):
+    def __init__(self, expr: pl.Expr, group_columns: list[str]) -> None:
         super().__init__(expr)
         self.group_columns = group_columns
 

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/collection.py

@@ -4,7 +4,7 @@
 import sys
 import warnings
 from abc import ABC
-from collections.abc import Mapping, MutableMapping, Sequence
+from collections.abc import Mapping, Sequence
 from pathlib import Path
 from typing import Any, Generic, Self, TypeVar, cast
 
@@ -20,10 +20,10 @@ from .random import Generator
 
 if sys.version_info >= (3, 13):
     SamplingType = TypeVar(
-        "SamplingType", bound=MutableMapping[str, Any], default=dict[str, Any]
+        "SamplingType", bound=Mapping[str, Any], default=Mapping[str, Any]
     )
 else:  # pragma: no cover
-    SamplingType = TypeVar("SamplingType", bound=MutableMapping[str, Any])
+    SamplingType = TypeVar("SamplingType", bound=Mapping[str, Any])
 
 
 class Collection(BaseCollection, ABC, Generic[SamplingType]):
@@ -123,7 +123,10 @@ class Collection(BaseCollection, ABC, Generic[SamplingType]):
                         ...
                     }
 
-                _Any_ member/value can be left out and will be sampled automatically.
+                *Any* member/value can be left out and will be sampled automatically.
+                Note that overrides for columns of members that are annotated with
+                ``inline_for_sampling=True`` can be supplied on the top-level instead
+                of in a nested dictionary.
             generator: The (seeded) generator to use for sampling data. If ``None``, a
                 generator with random seed is automatically created.
 
@@ -198,7 +201,11 @@ class Collection(BaseCollection, ABC, Generic[SamplingType]):
                             else _extract_keys_if_exist(sample, primary_keys)
                         ),
                         **_extract_keys_if_exist(
-                            sample[member] if member in sample else {},
+                            (
+                                sample
+                                if member_infos[member].inline_for_sampling
+                                else (sample[member] if member in sample else {})
+                            ),
                             schema.column_names(),
                         ),
                     }
@@ -498,7 +505,7 @@ class Collection(BaseCollection, ABC, Generic[SamplingType]):
 
     # ---------------------------------- PERSISTENCE --------------------------------- #
 
-    def write_parquet(self, directory: Path):
+    def write_parquet(self, directory: Path) -> None:
         """Write the members of this collection to Parquet files in a directory.
 
         This method writes one Parquet file per member into the provided directory.
@@ -590,7 +597,7 @@ class Collection(BaseCollection, ABC, Generic[SamplingType]):
         return out
 
     @classmethod
-    def _validate_input_keys(cls, data: Mapping[str, FrameType], /):
+    def _validate_input_keys(cls, data: Mapping[str, FrameType], /) -> None:
         actual = set(data)
 
         missing = cls.required_members() - actual

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/columns/__init__.py

@@ -10,6 +10,7 @@ from .enum import Enum
 from .float import Float, Float32, Float64
 from .integer import Int8, Int16, Int32, Int64, Integer, UInt8, UInt16, UInt32, UInt64
 from .list import List
+from .object import Object
 from .string import String
 from .struct import Struct
 
@@ -31,6 +32,7 @@ __all__ = [
     "Int32",
     "Int64",
     "Integer",
+    "Object",
     "UInt8",
     "UInt16",
     "UInt32",

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/columns/_mixins.py

@@ -83,7 +83,7 @@ U = TypeVar("U")
 class IsInMixin(Generic[U], Base):
     """Mixin to use for types implementing "is in"."""
 
-    def __init__(self, *, is_in: Sequence[U] | None = None, **kwargs: Any):
+    def __init__(self, *, is_in: Sequence[U] | None = None, **kwargs: Any) -> None:
         super().__init__(**kwargs)
         self.is_in = is_in
 

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/columns/decimal.py

@@ -148,7 +148,9 @@ class Decimal(OrdinalMixin[decimal.Decimal], Column):
 # --------------------------------------- UTILS -------------------------------------- #
 
 
-def _validate(value: decimal.Decimal, precision: int | None, scale: int, name: str):
+def _validate(
+    value: decimal.Decimal, precision: int | None, scale: int, name: str
+) -> None:
     exponent = value.as_tuple().exponent
     if not isinstance(exponent, int):
         raise ValueError(f"Encountered 'inf' or 'NaN' for `{name}`.")

dataframely-1.2.0/dataframely/columns/object.py

@@ -0,0 +1,63 @@
+# Copyright (c) QuantCo 2025-2025
+# SPDX-License-Identifier: BSD-3-Clause
+
+from __future__ import annotations
+
+from collections.abc import Callable
+from typing import Any
+
+import polars as pl
+
+from dataframely._compat import pa, sa, sa_TypeEngine
+from dataframely.random import Generator
+
+from ._base import Column
+
+
+class Object(Column):
+    """A Python Object column."""
+
+    def __init__(
+        self,
+        *,
+        nullable: bool = True,
+        primary_key: bool = False,
+        check: Callable[[pl.Expr], pl.Expr] | None = None,
+        alias: str | None = None,
+        metadata: dict[str, Any] | None = None,
+    ):
+        """
+        Args:
+            nullable: Whether this column may contain null values.
+            primary_key: Whether this column is part of the primary key of the schema.
+            check: A custom check to run for this column. Must return a non-aggregated
+                boolean expression.
+            alias: An overwrite for this column's name which allows for using a column
+                name that is not a valid Python identifier. Especially note that setting
+                this option does _not_ allow to refer to the column with two different
+                names, the specified alias is the only valid name.
+            metadata: A dictionary of metadata to attach to the column.
+        """
+        super().__init__(
+            nullable=nullable,
+            primary_key=primary_key,
+            check=check,
+            alias=alias,
+            metadata=metadata,
+        )
+
+    @property
+    def dtype(self) -> pl.DataType:
+        return pl.Object()
+
+    def sqlalchemy_dtype(self, dialect: sa.Dialect) -> sa_TypeEngine:
+        raise NotImplementedError("SQL column cannot have 'Object' type.")
+
+    @property
+    def pyarrow_dtype(self) -> pa.DataType:
+        raise NotImplementedError("PyArrow column cannot have 'Object' type.")
+
+    def _sample_unchecked(self, generator: Generator, n: int) -> pl.Series:
+        raise NotImplementedError(
+            "Random data sampling not implemented for 'Object' type."
+        )

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/config.py

@@ -25,23 +25,23 @@ class Config(contextlib.ContextDecorator):
     #: Singleton stack to track where to go back after exiting a context.
     _stack: list[Options] = []
 
-    def __init__(self, **options: Unpack[Options]):
+    def __init__(self, **options: Unpack[Options]) -> None:
         self._local_options: Options = {**default_options(), **options}
 
     @staticmethod
-    def set_max_sampling_iterations(iterations: int):
+    def set_max_sampling_iterations(iterations: int) -> None:
         """Set the maximum number of sampling iterations to use on
         :meth:`Schema.sample`."""
         Config.options["max_sampling_iterations"] = iterations
 
     @staticmethod
-    def restore_defaults():
+    def restore_defaults() -> None:
         """Restore the defaults of the configuration."""
         Config.options = default_options()
 
     # ------------------------------------ CONTEXT ----------------------------------- #
 
-    def __enter__(self):
+    def __enter__(self) -> None:
         Config._stack.append(Config.options)
         Config.options = self._local_options
 
@@ -50,5 +50,5 @@ class Config(contextlib.ContextDecorator):
         exc_type: type[BaseException] | None,
         exc_val: BaseException | None,
         exc_tb: TracebackType | None,
-    ):
+    ) -> None:
         Config.options = Config._stack.pop()

{dataframely-1.0.0 → dataframely-1.2.0}/dataframely/exc.py

@@ -11,7 +11,7 @@ from ._polars import PolarsDataType
 class ValidationError(Exception):
     """Error raised when :mod:`dataframely` validation encounters an issue."""
 
-    def __init__(self, message: str):
+    def __init__(self, message: str) -> None:
         super().__init__()
         self.message = message
 
@@ -22,7 +22,9 @@ class ValidationError(Exception):
 class DtypeValidationError(ValidationError):
     """Validation error raised when column dtypes are wrong."""
 
-    def __init__(self, errors: dict[str, tuple[PolarsDataType, PolarsDataType]]):
+    def __init__(
+        self, errors: dict[str, tuple[PolarsDataType, PolarsDataType]]
+    ) -> None:
         super().__init__(f"{len(errors)} columns have an invalid dtype")
         self.errors = errors
 
@@ -37,7 +39,7 @@ class DtypeValidationError(ValidationError):
 class RuleValidationError(ValidationError):
     """Complex validation error raised when rule validation fails."""
 
-    def __init__(self, errors: dict[str, int]):
+    def __init__(self, errors: dict[str, int]) -> None:
         super().__init__(f"{len(errors)} rules failed validation")
 
         # Split into schema errors and column errors
@@ -75,11 +77,11 @@ class RuleValidationError(ValidationError):
 class MemberValidationError(ValidationError):
     """Validation error raised when multiple members of a collection fail validation."""
 
-    def __init__(self, errors: dict[str, ValidationError]):
+    def __init__(self, errors: dict[str, ValidationError]) -> None:
         super().__init__(f"{len(errors)} members failed validation")
         self.errors = errors
 
-    def __str__(self):
+    def __str__(self) -> str:
         details = [
             f" > Member '{name}' failed validation:\n"
             + "\n".join("   " + line for line in str(error).split("\n"))
@@ -95,7 +97,7 @@ class ImplementationError(Exception):
 class AnnotationImplementationError(ImplementationError):
     """Error raised when the annotations of a collection are invalid."""
 
-    def __init__(self, attr: str, kls: type):
+    def __init__(self, attr: str, kls: type) -> None:
         message = (
             "Annotations of a 'dy.Collection' may only be an (optional) "
             f"'dy.LazyFrame', but \"{attr}\" has type '{kls}'."
@@ -106,7 +108,9 @@ class AnnotationImplementationError(ImplementationError):
 class RuleImplementationError(ImplementationError):
     """Error raised when a rule is implemented incorrectly."""
 
-    def __init__(self, name: str, return_dtype: pl.DataType, is_group_rule: bool):
+    def __init__(
+        self, name: str, return_dtype: pl.DataType, is_group_rule: bool
+    ) -> None:
         if is_group_rule:
             details = (
                 " When implementing a group rule (i.e. when using the `group_by` "