dataeval 1.1.3__tar.gz → 1.1.5__tar.gz

{dataeval-1.1.3 → dataeval-1.1.5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: dataeval
-Version: 1.1.3
+Version: 1.1.5
 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/

{dataeval-1.1.3 → dataeval-1.1.5}/pyproject.toml

@@ -107,10 +107,6 @@ lint = [
 docsync = [
   "jupytext>=1.19.1",
 ]
-docker = [
-  "jinja2>=3.1",
-  "pyyaml>=6.0",
-]
 test = [
   "coverage[toml]>=7.6",
   "filelock>=3.20.3",
@@ -318,6 +314,7 @@ markers = [
   "required: marks tests for required features",
   "optional: marks tests for optional features",
   "cuda: marks tests that require cuda",
+  "performance: marks verification tests that time workloads or measure memory",
 ]
 
 [tool.coverage.run]

{dataeval-1.1.3 → dataeval-1.1.5}/src/dataeval/_version.py

@@ -18,7 +18,7 @@ version_tuple: tuple[int | str, ...]
 commit_id: str | None
 __commit_id__: str | None
 
-__version__ = version = '1.1.3'
-__version_tuple__ = version_tuple = (1, 1, 3)
+__version__ = version = '1.1.5'
+__version_tuple__ = version_tuple = (1, 1, 5)
 
 __commit_id__ = commit_id = None

{dataeval-1.1.3 → dataeval-1.1.5}/src/dataeval/extractors/_bovw.py

@@ -208,7 +208,9 @@ class BoVWExtractor:
             If no SIFT features are found in any image. This typically occurs
             when all images are uniform (e.g., solid color).
         """
-        with PoolWrapper(get_max_processes(), context="spawn") as pool:
+        # Threads, not processes: OpenCV and k-means release the GIL, so threads scale SIFT
+        # without paying process startup and model pickling on every Embeddings batch.
+        with PoolWrapper(get_max_processes(), context="threads") as pool:
             all_descriptors = [
                 des for _, des in sorted(pool.imap_unordered(_extract_single, enumerate(data))) if len(des) > 0
             ]
@@ -269,7 +271,7 @@ class BoVWExtractor:
 
         result = np.zeros((len(images), n_clusters), dtype=np.float32)
 
-        with PoolWrapper(get_max_processes(), context="spawn") as pool:
+        with PoolWrapper(get_max_processes(), context="threads") as pool:
             for idx, histogram in pool.imap_unordered(
                 partial(_transform_single, kmeans=self._kmeans, n_clusters=n_clusters),
                 enumerate(images),

{dataeval-1.1.3 → dataeval-1.1.5}/src/dataeval/quality/_outliers.py

@@ -84,8 +84,8 @@ class OutliersOutput(DataFrameOutput, Generic[TOutliers]):
         Preserved across re-detection calls unless overridden via
         :meth:`with_threshold`.
     dataset_steps : list[int] or None
-        Cumulative dataset boundaries for multi-dataset index remapping.
-        None for single-dataset output.
+        Cumulative image counts per dataset, used to number cluster rows. Statistics rows are
+        numbered by steps recomputed from `calculation_results`. None for single-dataset output.
     """
 
     def __init__(
@@ -1752,7 +1752,8 @@ class Outliers(Evaluator):
         outliers_dfs: list[pl.DataFrame] = []
 
         if stats_results:
-            combined_stats, combined_source_index, _ = combine_stats_results(stats_results)
+            # combined_steps count statistics rows, target rows included; dataset_steps count images, for cluster rows.
+            combined_stats, combined_source_index, combined_steps = combine_stats_results(stats_results)
 
             class_ids: NDArray[np.intp] | None = None
             if per_class and metadata is not None:
@@ -1760,7 +1761,7 @@ class Outliers(Evaluator):
 
             outliers_df = self._get_outliers(combined_stats, combined_source_index, class_ids)
             outliers_df = drop_null_index_columns(outliers_df, ["target_index", "channel_index"])
-            outliers_df = add_dataset_index(outliers_df, dataset_steps)
+            outliers_df = add_dataset_index(outliers_df, combined_steps)
             outliers_dfs.append(outliers_df)
 
         # Cluster-based: early fusion — combine all images, cluster globally

{dataeval-1.1.3 → dataeval-1.1.5}/src/dataeval/utils/_internal.py

@@ -8,6 +8,7 @@ import sys
 from collections.abc import Callable, Iterable, Iterator, Mapping, Sequence
 from enum import Enum
 from importlib import import_module
+from multiprocessing.pool import ThreadPool
 from os import cpu_count
 from types import ModuleType
 from typing import Any, Literal, TypeVar, cast, overload
@@ -948,9 +949,14 @@ class PoolWrapper:
     Also supports 'threads' for workloads where the GIL is released during computation.
     """
 
-    def __init__(self, processes: int | None, context: Literal["fork", "spawn"] = DEFAULT_CONTEXT) -> None:
+    def __init__(self, processes: int | None, context: Literal["fork", "spawn", "threads"] = DEFAULT_CONTEXT) -> None:
         procs = 1 if processes is None else max(1, (cpu_count() or 1) + processes + 1) if processes < 0 else processes
-        self._pool = multiprocessing.get_context(context).Pool(procs) if procs > 1 else None
+        if procs <= 1:
+            self._pool = None
+        elif context == "threads":
+            self._pool = ThreadPool(procs)
+        else:
+            self._pool = multiprocessing.get_context(context).Pool(procs)
 
     def imap_unordered(self, func: Callable[[T], R], iterable: Iterable[T]) -> Iterator[R]:
         """Apply `func` to each item in `iterable`, optionally using a pool."""

{dataeval-1.1.3 → dataeval-1.1.5}/src/dataeval/utils/preprocessing.py

@@ -286,7 +286,7 @@ def to_bounding_box(boxlike: BoxLike, image_shape: tuple[int, ...] | None = None
         if isinstance(boxlike, Iterable):
             return BoundingBox(*boxlike, image_shape=image_shape)
         if isinstance(image_shape, tuple) and len(image_shape) > 2:
-            return BoundingBox(0, 0, image_shape[-2], image_shape[-1], image_shape=image_shape)
+            return BoundingBox(0, 0, image_shape[-1], image_shape[-2], image_shape=image_shape)
     except (TypeError, ValueError):
         _logger.warning(f"Invalid bounding box format: {boxlike}. Expected a BoundingBox or a tuple/list of 4 numbers.")