dataeval 1.0.6__tar.gz → 1.1.0rc0__tar.gz

{dataeval-1.0.6 → dataeval-1.1.0rc0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: dataeval
-Version: 1.0.6
+Version: 1.1.0rc0
 Summary: DataEval provides a simple interface to characterize image data and its impact on model performance across classification and object-detection tasks
 Project-URL: Homepage, https://dataeval.ai/
 Project-URL: Repository, https://github.com/aria-ml/dataeval/
@@ -21,6 +21,7 @@ Classifier: Programming Language :: Python :: 3.14
 Classifier: Topic :: Scientific/Engineering
 Requires-Python: >=3.10
 Requires-Dist: lightgbm>=4
+Requires-Dist: maite>=0.9.4
 Requires-Dist: numba>=0.61.0
 Requires-Dist: numpy>=1.24.2
 Requires-Dist: polars>=1.0.0
@@ -37,23 +38,27 @@ Requires-Dist: torchvision>=0.17.0; extra == 'cpu'
 Provides-Extra: cu118
 Requires-Dist: torch>=2.2.0; extra == 'cu118'
 Requires-Dist: torchvision>=0.17.0; extra == 'cu118'
-Provides-Extra: cu124
-Requires-Dist: torch>=2.2.0; extra == 'cu124'
-Requires-Dist: torchvision>=0.17.0; extra == 'cu124'
 Provides-Extra: cu128
 Requires-Dist: torch>=2.2.0; extra == 'cu128'
 Requires-Dist: torchvision>=0.17.0; extra == 'cu128'
+Provides-Extra: litert
+Requires-Dist: ai-edge-litert>=2.0; (python_version <= '3.14') and extra == 'litert'
 Provides-Extra: onnx
-Requires-Dist: onnx; extra == 'onnx'
-Requires-Dist: onnxruntime>=1.14.0; extra == 'onnx'
+Requires-Dist: onnx>=1.14.0; extra == 'onnx'
+Requires-Dist: onnxruntime<1.24,>=1.14.0; (python_version == '3.10') and extra == 'onnx'
+Requires-Dist: onnxruntime>=1.14.0; (python_version >= '3.11') and extra == 'onnx'
 Provides-Extra: onnx-gpu
-Requires-Dist: onnx; extra == 'onnx-gpu'
-Requires-Dist: onnxruntime-gpu>=1.14.0; extra == 'onnx-gpu'
+Requires-Dist: onnx>=1.14.0; extra == 'onnx-gpu'
+Requires-Dist: onnxruntime-gpu<1.24,>=1.14.0; (python_version == '3.10') and extra == 'onnx-gpu'
+Requires-Dist: onnxruntime-gpu>=1.14.0; (python_version >= '3.11') and extra == 'onnx-gpu'
+Provides-Extra: ontology
+Requires-Dist: rdflib>=7.0; extra == 'ontology'
 Provides-Extra: opencv
 Requires-Dist: opencv-python-headless>=4.8.0; extra == 'opencv'
 Description-Content-Type: text/markdown
 
 <!-- markdownlint-disable MD041 -->
+
 ![dataeval-logo](docs/source/_static/images/DataEval_ImageText.png)
 
 <!-- :auto badges: -->
@@ -130,14 +135,28 @@ You can install DataEval directly from pypi.org using the following command.
 pip install dataeval
 ```
 
+By default, PyTorch is installed from PyPI which includes CUDA support on Linux.
+To install a specific PyTorch variant, use `--extra-index-url`:
+
+```bash
+# CPU only
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cpu
+
+# CUDA 11.8
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cu118
+
+# CUDA 12.8
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cu128
+```
+
 ### **Installing with conda**
 
 DataEval can be installed in a Conda/Mamba environment using the provided
-`environment.yaml` file. As some dependencies are installed from the `pytorch`
+`environment.yml` file. As some dependencies are installed from the `pytorch`
 channel, the channel is specified in the below example.
 
 ```bash
-micromamba create -f environment\environment.yaml -c pytorch
+micromamba create -f environment\environment.yml -c pytorch
 ```
 
 ### **Installing from GitHub**
@@ -401,7 +420,7 @@ shape: (3, 5)
 
 A result with many large groups is a signal that your dataset contains
 repeated collection events. Before training, remove all but one sample from
-each group. See the [deduplication how-to guide](./docs/source/notebooks/h2_deduplicate.md)
+each group. See the [deduplication how-to guide](./docs/source/notebooks/h2_deduplicate.py)
 for a complete walkthrough, including how to choose which sample to keep.
 
 ### Where to go next

{dataeval-1.0.6 → dataeval-1.1.0rc0}/README.md

@@ -1,4 +1,5 @@
 <!-- markdownlint-disable MD041 -->
+
 ![dataeval-logo](docs/source/_static/images/DataEval_ImageText.png)
 
 <!-- :auto badges: -->
@@ -75,14 +76,28 @@ You can install DataEval directly from pypi.org using the following command.
 pip install dataeval
 ```
 
+By default, PyTorch is installed from PyPI which includes CUDA support on Linux.
+To install a specific PyTorch variant, use `--extra-index-url`:
+
+```bash
+# CPU only
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cpu
+
+# CUDA 11.8
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cu118
+
+# CUDA 12.8
+pip install dataeval --extra-index-url https://download.pytorch.org/whl/cu128
+```
+
 ### **Installing with conda**
 
 DataEval can be installed in a Conda/Mamba environment using the provided
-`environment.yaml` file. As some dependencies are installed from the `pytorch`
+`environment.yml` file. As some dependencies are installed from the `pytorch`
 channel, the channel is specified in the below example.
 
 ```bash
-micromamba create -f environment\environment.yaml -c pytorch
+micromamba create -f environment\environment.yml -c pytorch
 ```
 
 ### **Installing from GitHub**
@@ -346,7 +361,7 @@ shape: (3, 5)
 
 A result with many large groups is a signal that your dataset contains
 repeated collection events. Before training, remove all but one sample from
-each group. See the [deduplication how-to guide](./docs/source/notebooks/h2_deduplicate.md)
+each group. See the [deduplication how-to guide](./docs/source/notebooks/h2_deduplicate.py)
 for a complete walkthrough, including how to choose which sample to keep.
 
 ### Where to go next

{dataeval-1.0.6 → dataeval-1.1.0rc0}/pyproject.toml

@@ -31,6 +31,7 @@ classifiers = [
   "Topic :: Scientific/Engineering",
 ]
 dependencies = [
+  "maite>=0.9.4",
   "numba>=0.61.0",
   "lightgbm>=4",
   "numpy>=1.24.2",
@@ -47,17 +48,43 @@ dependencies = [
 [project.optional-dependencies]
 cpu = ["torch>=2.2.0", "torchvision>=0.17.0"]
 cu118 = ["torch>=2.2.0", "torchvision>=0.17.0"]
-cu124 = ["torch>=2.2.0", "torchvision>=0.17.0"]
 cu128 = ["torch>=2.2.0", "torchvision>=0.17.0"]
+litert = ["ai-edge-litert>=2.0; python_version <= '3.14'"]
 opencv = ["opencv-python-headless>=4.8.0"]
-onnx = ["onnx", "onnxruntime>=1.14.0"]
-onnx-gpu = ["onnx", "onnxruntime-gpu>=1.14.0"]
+onnx = [
+  "onnx>=1.14.0",
+  "onnxruntime>=1.14.0,<1.24; python_version == '3.10'",
+  "onnxruntime>=1.14.0; python_version >= '3.11'",
+]
+onnx-gpu = [
+  "onnx>=1.14.0",
+  "onnxruntime-gpu>=1.14.0,<1.24; python_version == '3.10'",
+  "onnxruntime-gpu>=1.14.0; python_version >= '3.11'",
+]
+ontology = ["rdflib>=7.0"]
 
 [project.urls]
 Homepage = "https://dataeval.ai/"
 Repository = "https://github.com/aria-ml/dataeval/"
 Documentation = "https://dataeval.readthedocs.io/"
 
+# MAITE interoperability entry-points.
+[project.entry-points."maite.tasks"]
+dataeval_balance = "dataeval.bias:Balance"
+dataeval_diversity = "dataeval.bias:Diversity"
+dataeval_parity = "dataeval.bias:Parity"
+dataeval_outliers = "dataeval.quality:Outliers"
+dataeval_duplicates = "dataeval.quality:Duplicates"
+dataeval_sufficiency = "dataeval.performance:Sufficiency"
+
+[project.entry-points."maite.protocols.image_classification.Model"]
+dataeval_OnnxImageClassifier = "dataeval.models:OnnxImageClassifier"
+dataeval_LiteRtImageClassifier = "dataeval.models:LiteRtImageClassifier"
+
+[project.entry-points."maite.protocols.object_detection.Model"]
+dataeval_OnnxObjectDetector = "dataeval.models:OnnxObjectDetector"
+dataeval_LiteRtObjectDetector = "dataeval.models:LiteRtObjectDetector"
+
 [dependency-groups]
 base = [
   "uv>=0.8.0",
@@ -65,7 +92,7 @@ base = [
 lock = [
   { include-group = "base" },
   "pyproject2conda>=0.22",
-  "poetry>=2; python_version<'3.14'",
+  "poetry==2.2.0; python_version<'3.14'",
 ]
 lint = [
   "ruff>=0.11",
@@ -73,19 +100,15 @@ lint = [
 ]
 docsync = [
   "jupytext>=1.19.1",
-  "mdformat-myst",
-]
-doclint = [
-  { include-group = "docs"},
-  "ruff>=0.11",
-  "pyright[nodejs]>=1.1.400",
 ]
 test = [
   "coverage[toml]>=7.6",
   "onnx>=1.14.0",
+  "onnxscript>=0.6.0",
   "pytest>=8.3",
   "pytest-cov>=6.1",
   "pytest-xdist>=3.6.1",
+  "rdflib>=7.0",
 ]
 verify = [
   "pytest>=8.3",
@@ -103,10 +126,13 @@ docs = [
   "jinja2>=3.1.6",
   "jupyter-client>=8.6.0",
   "jupyter-cache>=1.0",
-  "maite-datasets>=0.0.12",
+  "maite-datasets>=0.0.15",
   "myst-nb>=1.0",
   "opencv-python-headless>=4.8.0",
+  "pandas>=2.0.0",
   "plotly>=6.2.0",
+  "rapidfuzz>=3.0",
+  "rdflib>=7.0",
   "sphinx-autoapi>=3.6.0",
   "sphinx-design>=0.6.1",
   "sphinx-immaterial>=0.12.5",
@@ -116,21 +142,15 @@ docs = [
   "torchmetrics>=1.0.0",
   "markupsafe>=3,<3.0.2",
   "jupytext>=1.19.1",
+  "ultralytics>=8.0.0",
 ]
 security = [  # keep in sync with [tool.uv.constraint-dependencies]
   "cryptography>=46.0.5",    # CVE-2026-26007: Missing Subgroup Validation for SECT Curves
   "filelock>=3.20.3",        # GHSA-w853-jp5j-5j7f, GHSA-qmgc-5h2g-mvrw
-  "onnx>=1.21.0",            # CVE-2026-28500: Untrusted Model Repository Warnings Suppressed by silent=True
-                             # CVE-2026-34445: Malicious ONNX models can crash servers by exploiting unprotected object settings
-                             # CVE-2026-27489: Vulnerable to Path Traversal via Symlink
-                             # GHSA-q56x-g2fj-4rj6: TOCTOU arbitrary file read/write in save_external_dat
   "pillow>=12.1.1",          # CVE-2026-25990: OOB write via PSD image
-  "poetry>=2.3.3",           # CVE-2026-34591: Poetry Has Wheel Path Traversal Which Can Lead to Arbitrary File Write
   "protobuf>=6.33.5",        # GHSA-7gcm-g887-7qv7
   "setuptools>=82.0.0",      # CVE-2026-23949: (jaraco_context) path traversal in tarball()
                              # CVE-2026-24049: (wheel) privilege escalation via unpack
-  "tornado>=6.5.5",          # CVE-2026-31958: Tornado is vulnerable to DoS due to too many multipart parts
-                             # CVE-2026-35536: Tornado has cookie attribute injection via .RequestHandler.set_cookie
 ]
 dev = [
   { include-group = "base" },
@@ -149,24 +169,16 @@ conflicts = [
   [
     { extra = "cpu" },
     { extra = "cu118" },
-    { extra = "cu124" },
     { extra = "cu128" },
   ],
 ]
 constraint-dependencies = [
     "cryptography>=46.0.5",    # CVE-2026-26007: Missing Subgroup Validation for SECT Curves
     "filelock>=3.20.3",        # GHSA-w853-jp5j-5j7f, GHSA-qmgc-5h2g-mvrw
-    "onnx>=1.21.0",            # CVE-2026-28500: Untrusted Model Repository Warnings Suppressed by silent=True
-                               # CVE-2026-34445: Malicious ONNX models can crash servers by exploiting unprotected object settings
-                               # CVE-2026-27489: Vulnerable to Path Traversal via Symlink
-                               # GHSA-q56x-g2fj-4rj6: TOCTOU arbitrary file read/write in save_external_dat
     "pillow>=12.1.1",          # CVE-2026-25990: OOB write via PSD image
-    "poetry>=2.3.3",           # CVE-2026-34591: Poetry Has Wheel Path Traversal Which Can Lead to Arbitrary File Write
     "protobuf>=6.33.5",        # GHSA-7gcm-g887-7qv7
     "setuptools>=82.0.0",      # CVE-2026-23949: (jaraco_context) path traversal in tarball()
                                # CVE-2026-24049: (wheel) privilege escalation via unpack
-    "tornado>=6.5.5",          # CVE-2026-31958: Tornado is vulnerable to DoS due to too many multipart parts
-                               # CVE-2026-35536: Tornado has cookie attribute injection via .RequestHandler.set_cookie
 ]
 
 [[tool.uv.index]]
@@ -179,11 +191,6 @@ name = "pytorch-cu118"
 url = "https://download.pytorch.org/whl/cu118"
 explicit = true
 
-[[tool.uv.index]]
-name = "pytorch-cu124"
-url = "https://download.pytorch.org/whl/cu124"
-explicit = true
-
 [[tool.uv.index]]
 name = "pytorch-cu128"
 url = "https://download.pytorch.org/whl/cu128"
@@ -193,19 +200,28 @@ explicit = true
 torch = [
   { index = "pytorch-cpu", extra = "cpu" },
   { index = "pytorch-cu118", extra = "cu118" },
-  { index = "pytorch-cu124", extra = "cu124" },
   { index = "pytorch-cu128", extra = "cu128" },
 ]
 torchvision = [
   { index = "pytorch-cpu", extra = "cpu" },
   { index = "pytorch-cu118", extra = "cu118" },
-  { index = "pytorch-cu124", extra = "cu124" },
   { index = "pytorch-cu128", extra = "cu128" },
 ]
 
 [tool.uv.extra-build-dependencies]
 numba = ["tbb>=2021.6"]
 
+[tool.poetry]
+version = "0.0.0"  # overwritten by poetry-dynamic-versioning
+
+[[tool.poetry.source]]
+name = "pytorch-cpu"
+url = "https://download.pytorch.org/whl/cpu"
+priority = "supplemental"
+
+[tool.poetry.dependencies]
+torch = { version = ">=2.2.0", source = "pytorch-cpu" }
+
 [tool.hatch.build.targets.sdist]
 include = ["src/dataeval"]
 
@@ -221,11 +237,11 @@ source = "vcs"
 [tool.hatch.build.hooks.vcs]
 version-file = "src/dataeval/_version.py"
 
-[tool.poetry]
-version = "0.0.0"  # unused
-
-[tool.poetry.dependencies]
-python = ">=3.10,<3.15"
+[tool.poetry-dynamic-versioning]
+enable = true
+vcs = "git"
+style = "pep440"
+pattern = "^v?(?P<base>\\d+\\.\\d+\\.\\d+)"
 
 [tool.pyproject2conda.dependencies]
 numpy = { skip = true, packages = "numpy>=1.24.2" }
@@ -235,7 +251,7 @@ torch = { pip = true }  # PyTorch is no longer maintained on conda-forge
 xxhash = { skip = true, packages = "python-xxhash>=3.3" }
 
 [tool.pyright]
-include = ["src", "tests"]
+include = ["src", "tests", "verification", "docs/source/notebooks"]
 exclude = [
   "**/__pycache__",
   "**/node_modules",
@@ -248,6 +264,10 @@ reportMissingImports = false
 
 [tool.pytest.ini_options]
 testpaths = ["tests"]
+filterwarnings = [
+  "ignore:The default value of normalize_pixel_values changed:FutureWarning",
+  "ignore:Clustering metrics expect discrete values but received continuous values:UserWarning",
+]
 addopts = [
   "--pythonwarnings=ignore::DeprecationWarning",
   "--verbose",
@@ -294,6 +314,7 @@ exclude = [
   ".tox",
   "prototype",
   "src/dataeval/_version.py",
+  "*.ipynb",
 ]
 line-length = 120
 indent-width = 4
@@ -308,7 +329,7 @@ ignore = ["ANN101", "ANN102", "ANN401", "C408", "C416", "COM812", "NPY002", "SLF
 fixable = ["ALL"]
 unfixable = []
 dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"
-per-file-ignores = { "!src/*" = ["ANN", "ARG", "BLE", "S", "SLF", "RET", "C90", "D", "FIX", "N", "PERF"], "docs/*" = ["B904"] }
+per-file-ignores = { "!src/*" = ["ANN", "ARG", "BLE", "S", "SLF", "RET", "C90", "D", "FIX", "N", "PERF"], "docs/*" = ["B904"], "docs/source/notebooks/*" = ["E402", "E501", "E703", "RUF100", "SIM105", "UP009"] }
 
 [tool.ruff.lint.flake8-builtins]
 builtins-strict-checking = false
@@ -332,8 +353,8 @@ docstring-code-format = true
 docstring-code-line-length = "dynamic"
 
 [tool.codespell]
-skip = './*env*,./output,./docs/build,./docs/source/.jupyter_cache,./docs/source/*/data,CHANGELOG.md,uv.lock,requirements.txt,*.html,*.lock,*.ipynb'
-ignore-words-list = ["Hart","FPR"]
+skip = './*env*,./output,./docs/build,./docs/source/.jupyter_cache,./docs/source/*/data,CHANGELOG.md,uv.lock,requirements.*.txt,*.html,*.lock,*.ipynb'
+ignore-words-list = ["Hart","FPR", "MOT", "mot"]
 
 [build-system]
 requires = ["hatchling", "hatch-vcs"]

{dataeval-1.0.6 → dataeval-1.1.0rc0}/src/dataeval/__init__.py

@@ -20,24 +20,30 @@ __all__ = [
     "exceptions",
     "flags",
     "log",
+    "models",
     "protocols",
     "types",
     "Embeddings",
     "Metadata",
+    "Ontology",
 ]
 
 import logging
 
-from . import config, exceptions, flags, protocols, types
+from . import config, exceptions, flags, models, protocols, types
 from ._embeddings import Embeddings
 from ._metadata import Metadata
+from ._ontology import Ontology
 
 logging.getLogger(__name__).addHandler(logging.NullHandler())
 
 
 def log(level: int = logging.DEBUG, handler: logging.Handler | None = None) -> None:
     """
-    Add a StreamHandler to the logger quickly for debugging.
+    Add a handler to the logger quickly for debugging.
+
+    Calling this more than once is idempotent: a handler equal to one already
+    attached to the logger is not added again, so log lines are not duplicated.
 
     Parameters
     ----------
@@ -45,18 +51,21 @@ def log(level: int = logging.DEBUG, handler: logging.Handler | None = None) -> N
         Set the logging level for the logger.
     handler : logging.Handler, optional
         Sets the logging handler for the logger if provided, otherwise logger will be
-        provided with a StreamHandler.
+        provided with a StreamHandler. When a custom handler is supplied its formatter
+        is left untouched; the default StreamHandler is given a verbose debugging
+        formatter.
     """
     import logging
 
-    logger = logging.getLogger(__name__)
+    _logger = logging.getLogger(__name__)
     if handler is None:
-        handler = logging.StreamHandler() if handler is None else handler
+        handler = logging.StreamHandler()
         handler.setFormatter(
             logging.Formatter(
                 "%(asctime)s %(levelname)-8s %(name)s.%(filename)s:%(lineno)s - %(funcName)10s() | %(message)s",
             ),
         )
-    logger.addHandler(handler)
-    logger.setLevel(level)
-    logger.debug(f"Added logging handler {handler} to logger: {__name__}")
+    if handler not in _logger.handlers:
+        _logger.addHandler(handler)
+    _logger.setLevel(level)
+    _logger.debug("Added logging handler %s to logger: %s", handler, __name__)

{dataeval-1.0.6 → dataeval-1.1.0rc0}/src/dataeval/_embeddings.py

@@ -1,6 +1,6 @@
 """Embeddings class for extracting and managing image embeddings."""
 
-__all__ = []
+__all__ = ["Embeddings"]
 
 import logging
 import os
@@ -14,7 +14,7 @@ import xxhash as xxh
 from numpy.typing import NDArray
 from typing_extensions import Self
 
-from dataeval.config import get_batch_size
+from dataeval.config import resolve_batch_size
 from dataeval.exceptions import NotFittedError
 from dataeval.extractors import FlattenExtractor
 from dataeval.protocols import (
@@ -25,6 +25,8 @@ from dataeval.protocols import (
     FeatureExtractor,
     ProgressCallback,
 )
+from dataeval.utils._internal import unwrap_image
+from dataeval.utils._validate import requires_maite_dataset
 
 _logger = logging.getLogger(__name__)
 
@@ -53,8 +55,14 @@ class Embeddings(Array, FeatureExtractor):
         :class:`~dataeval.extractors.FlattenExtractor` for simple baseline
         compatibility with all DataEval tools.
     batch_size : int or None, default None
-        Number of samples to process per batch. When None, uses DataEval's
-        configured batch size via :func:`~dataeval.config.get_batch_size`.
+        I/O chunk size: how many images are loaded from the dataset, encoded, and
+        written to storage per step. Resolved via
+        :func:`~dataeval.config.resolve_batch_size` as the first set of
+        ``batch_size`` (this argument), the extractor's own ``batch_size``, then
+        the global default. This is distinct from an extractor's own forward-pass
+        (compute) batch size: an extractor with its own ``batch_size`` sub-batches
+        each chunk for the model, so the smaller of the two bounds the forward
+        pass. Batching never changes the resulting embeddings.
     path : Path, str, or None, default None
         File path for memory-mapped storage. When None, caches embeddings in memory only.
         When Path or string is provided, uses memory-mapped storage for large embeddings
@@ -93,6 +101,7 @@ class Embeddings(Array, FeatureExtractor):
 
     memory_threshold: float
 
+    @requires_maite_dataset("dataset", expected="image_only")
     def __init__(
         self,
         # Technically more permissive than ImageClassificationDataset or ObjectDetectionDataset
@@ -104,7 +113,7 @@ class Embeddings(Array, FeatureExtractor):
         progress_callback: ProgressCallback | None = None,
     ) -> None:
         self._extractor = extractor if extractor is not None else FlattenExtractor()
-        self._batch_size = get_batch_size(batch_size)
+        self._batch_size = resolve_batch_size(batch_size, getattr(self._extractor, "batch_size", None))
         self.memory_threshold = max(0.0, min(1.0, memory_threshold))
         self._progress_callback = progress_callback
 
@@ -159,6 +168,7 @@ class Embeddings(Array, FeatureExtractor):
         """
         return self._dataset is not None
 
+    @requires_maite_dataset("dataset", expected="image_only")
     def bind(self, dataset: Dataset[tuple[ArrayLike, Any, Any]] | Dataset[ArrayLike]) -> Self:
         """Bind this instance to a dataset.
 
@@ -502,12 +512,7 @@ class Embeddings(Array, FeatureExtractor):
         if self._dataset is None:
             raise NotFittedError("No dataset bound. Call bind() first.")
 
-        images: list[Any] = []
-        for idx in indices:
-            item = self._dataset[idx]
-            image = item[0] if isinstance(item, tuple) else item
-            images.append(image)
-        return images
+        return [unwrap_image(self._dataset[idx]) for idx in indices]
 
     def _batch(self, indices: Sequence[int]) -> Iterator[NDArray[Any]]:  # noqa: C901
         """Process indices in batches using the extractor."""

{dataeval-1.0.6 → dataeval-1.1.0rc0}/src/dataeval/_experimental.py

@@ -39,14 +39,6 @@ def _make_warning_message(  # noqa: C901
     return msg
 
 
-def _prepend_doc_note(doc: str | None, note: str) -> str:
-    """Prepend a status note to a docstring."""
-    header = f".. warning::\n    {note}"
-    if doc:
-        return f"{header}\n\n{doc}"
-    return header
-
-
 @overload
 def experimental(_target: F) -> F: ...
 @overload
@@ -89,7 +81,7 @@ def experimental(  # noqa: C901
                 original_init(self, *args, **kwargs)
 
             target.__init__ = new_init  # type: ignore[attr-defined]
-            target.__doc__ = _prepend_doc_note(target.__doc__, msg)
+            target.__experimental__ = msg  # type: ignore[attr-defined]
             return target  # type: ignore[return-value]
 
         @functools.wraps(target)
@@ -100,7 +92,7 @@ def experimental(  # noqa: C901
                 warned = True
             return target(*args, **kwargs)
 
-        wrapper.__doc__ = _prepend_doc_note(target.__doc__, msg)
+        wrapper.__experimental__ = msg  # type: ignore[attr-defined]
         return wrapper  # type: ignore[return-value]
 
     if _target is not None:
@@ -165,7 +157,7 @@ def deprecated(  # noqa: C901
                 original_init(self, *args, **kwargs)
 
             target.__init__ = new_init  # type: ignore[attr-defined]
-            target.__doc__ = _prepend_doc_note(target.__doc__, msg)
+            target.__deprecated__ = msg  # type: ignore[attr-defined]
             return target  # type: ignore[return-value]
 
         @functools.wraps(target)
@@ -176,7 +168,7 @@ def deprecated(  # noqa: C901
                 warned = True
             return target(*args, **kwargs)
 
-        wrapper.__doc__ = _prepend_doc_note(target.__doc__, msg)
+        wrapper.__deprecated__ = msg  # type: ignore[attr-defined]
         return wrapper  # type: ignore[return-value]
 
     if _target is not None:

{dataeval-1.0.6 → dataeval-1.1.0rc0}/src/dataeval/_metadata.py

@@ -1,4 +1,4 @@
-__all__ = []
+__all__ = ["Metadata"]
 
 import logging
 from collections.abc import Callable, Iterable, Iterator, Mapping, Sequence, Sized
@@ -22,6 +22,7 @@ from dataeval.protocols import (
 )
 from dataeval.types import Array1D
 from dataeval.utils._internal import as_numpy, merge_metadata
+from dataeval.utils._validate import requires_maite_dataset
 
 _logger = logging.getLogger(__name__)
 
@@ -105,6 +106,7 @@ class Metadata(Array, FeatureExtractor):
     >>> test_factors = metadata(test_dataset)  # Extract from new dataset
     """
 
+    @requires_maite_dataset("dataset", expected="any_target")
     def __init__(
         self,
         dataset: AnnotatedDataset[tuple[Any, Any, DatumMetadata]] | None = None,
@@ -168,6 +170,7 @@ class Metadata(Array, FeatureExtractor):
         """
         return self._dataset is not None
 
+    @requires_maite_dataset("dataset", expected="any_target")
     def bind(self, dataset: AnnotatedDataset[tuple[Any, Any, DatumMetadata]]) -> Self:
         """Bind this instance to a dataset.
 
@@ -573,6 +576,11 @@ class Metadata(Array, FeatureExtractor):
             Rows where target_index is None contain datum-level data.
             Rows where target_index is an integer contain target/detection-level data.
 
+        See Also
+        --------
+        :attr:`~dataeval.Metadata.image_data` : Filter to image-level rows only
+        :attr:`~dataeval.Metadata.target_data` : Filter to target-level rows only
+
         Notes
         -----
         This property triggers dataset structure analysis on first access.
@@ -581,11 +589,6 @@ class Metadata(Array, FeatureExtractor):
         For Object Detection datasets, the dataframe now contains:
         - Image-level rows (target_index=None): One per image with image-level factors
         - Target-level rows (target_index=0,1,2...): One per detection with detection data
-
-        See Also
-        --------
-        image_data : Filter to image-level rows only
-        target_data : Filter to target-level rows only
         """
         self._structure()
         return self._dataframe