Merge branch 'sort-fix' into embedding-fix

README.md

@@ -83,6 +83,7 @@ For more details on setting up a development environment see [DEVELOPMENT docs](
 1. Merge `main` to `release` with a _merge commit_.
 2. Run "Create Release" workflow with workflow from `release` branch.
 3. Merge `release` to `main` with a _merge commit_.
+4. Check package versions in Conda Feedstock [meta.yaml file](https://github.com/conda-forge/nrtk-explorer-feedstock/blob/main/recipe/meta.yaml)
 
 [1]: https://trame.readthedocs.io/en/latest/
 [2]: https://www.kitware.com/

pyproject.toml

@@ -30,7 +30,7 @@ dependencies = [
     "accelerate",
     "numpy",
     "Pillow",
-    "pybsm>=0.6,<=0.9.0",
+    "pybsm==0.10.2",
     "scikit-learn>=1.6.0",
     "smqtk_image_io",
     "tabulate",
@@ -42,15 +42,12 @@ dependencies = [
     "transformers",
     "datasets[vision]",
     "umap-learn",
-    "nrtk[headless]>=0.12.0,<=0.16.0",
+    "nrtk[headless]==0.19.1",
     "trame-annotations>=0.4.0",
+    "kwcoco",
 ]
 
 [project.optional-dependencies]
-kwcoco= [
-  "kwcoco",
-]
-
 dev = [
   "black",
   "flake8",

src/nrtk_explorer/app/embeddings.py

@@ -5,6 +5,7 @@
 from nrtk_explorer.library import embeddings_extractor
 from nrtk_explorer.library import dimension_reducers
 from nrtk_explorer.library.dataset import get_dataset
+from nrtk_explorer.library.scoring import partition
 from nrtk_explorer.app.applet import Applet
 
 from nrtk_explorer.app.images.image_ids import (
@@ -230,7 +231,6 @@ def update_transformed_images(self, id_to_image):
             **self._stashed_points_transformations,
             **updated_points,
         }
-
         self.update_points_transformations_state()
 
     # called by category filter

src/nrtk_explorer/app/transforms.py

@@ -9,7 +9,6 @@
 from trame_server import Server
 
 import nrtk_explorer.library.transforms as trans
-import nrtk_explorer.library.nrtk_transforms as nrtk_trans
 import nrtk_explorer.library.yaml_transforms as nrtk_yaml
 from nrtk_explorer.library.multiprocess_predictor import MultiprocessPredictor
 from nrtk_explorer.library.app_config import process_config
@@ -204,9 +203,6 @@ def delete_meta_state(old_ids, new_ids):
             "identity": trans.IdentityTransform,
         }
 
-        if nrtk_trans.nrtk_transforms_available():
-            self._transform_classes["nrtk_pybsm"] = nrtk_trans.NrtkPybsmTransform
-
         # Add transform from YAML definition
         self._transform_classes.update(nrtk_yaml.generate_transforms())
 

src/nrtk_explorer/library/annotations.py

@@ -1,5 +1,5 @@
 from typing import TypedDict, List
-from .dataset import JsonDataset
+from .dataset import CocoDataset
 
 
 def get_cat_id(dataset, annotation):
@@ -33,7 +33,7 @@ class Annotation(TypedDict, total=False):
     bbox: List[float]
 
 
-def to_annotation(dataset: JsonDataset, prediction: Prediction) -> Annotation:
+def to_annotation(dataset: CocoDataset, prediction: Prediction) -> Annotation:
     annotation: Annotation = {}
 
     if "label" in prediction:

src/nrtk_explorer/library/dataset.py

@@ -10,8 +10,8 @@
 import os
 from functools import lru_cache
 from pathlib import Path
-import json
 from PIL import Image
+import kwcoco
 from datasets import (
     load_dataset,
     get_dataset_infos,
@@ -36,42 +36,12 @@ def build_cat_index(self):
         self.name_to_cat = {cat["name"]: cat for cat in self.cats.values()}
 
 
-class JsonDataset(BaseDataset, CategoryIndex):
-    """JSON-based COCO datasets."""
-
-    def __init__(self, path: str):
-        with open(path) as f:
-            self.data = json.load(f)
-        self.fpath = path
-        self.cats = {cat["id"]: cat for cat in self.data["categories"]}
-        self.anns = {ann["id"]: ann for ann in self.data["annotations"]}
-        self.imgs = {img["id"]: img for img in self.data["images"]}
-        self.build_cat_index()
-
-    def _get_image_fpath(self, selected_id: int):
-        dataset_dir = Path(self.fpath).parent
-        file_name = self.imgs[selected_id]["file_name"]
-        return str(dataset_dir / file_name)
-
+class CocoDataset(kwcoco.CocoDataset, BaseDataset):
     def get_image(self, id: int):
-        image_fpath = self._get_image_fpath(id)
+        image_fpath = self.get_image_fpath(id)
         return Image.open(image_fpath)
 
 
-def make_coco_dataset(path: str):
-    try:
-        import kwcoco
-
-        class CocoDataset(kwcoco.CocoDataset, BaseDataset):
-            def get_image(self, id: int):
-                image_fpath = self.get_image_fpath(id)
-                return Image.open(image_fpath)
-
-        return CocoDataset(path)
-    except ImportError:
-        return JsonDataset(path)
-
-
 def is_coco_dataset(path: str):
     if not os.path.exists(path) or os.path.isdir(path):
         return False
@@ -102,7 +72,7 @@ def find_column_name(features, column_names):
 
 
 class HuggingFaceDataset(BaseDataset, CategoryIndex):
-    """Interface for Hugging Face datasets with a similar API to JsonDataset."""
+    """Interface for Hugging Face datasets with a similar API to CocoDataset."""
 
     def __init__(self, identifier: str):
         self.imgs: dict[str, dict] = {}
@@ -244,7 +214,7 @@ def get_dataset(identifier: str):
     absolute_path = str(Path(identifier).resolve())
 
     if is_coco_dataset(absolute_path):
-        return make_coco_dataset(absolute_path)
+        return CocoDataset(absolute_path)
 
     # Assume identifier is a Hugging Face Dataset
     return HuggingFaceDataset(identifier)

src/nrtk_explorer/library/nrtk_transforms.py

@@ -1,33 +1,14 @@
-from typing import Any, Dict, Optional, TYPE_CHECKING
-
 import numpy as np
 import logging
-from PIL import Image as ImageModule
-from PIL.Image import Image
-from nrtk_explorer.library.transforms import ImageTransform, ParameterDescription
-
-ENABLED_NRTK_TRANSFORMS = True
 
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.INFO)
 
 try:
     from pybsm.otf import dark_current_from_density
-    from nrtk.impls.perturb_image.pybsm.perturber import PybsmPerturber, PybsmSensor, PybsmScenario
+    from nrtk.impls.perturb_image.pybsm.perturber import PybsmSensor, PybsmScenario
 except ImportError:
     logger.info("Disabling NRTK transforms due to missing library/failing imports")
-    ENABLED_NRTK_TRANSFORMS = False
-
-if TYPE_CHECKING:
-    PybsmPerturberType = PybsmPerturber
-else:
-    PybsmPerturberType = None
-
-PybsmPerturberArg = Optional[PybsmPerturberType]
-
-
-def nrtk_transforms_available():
-    return ENABLED_NRTK_TRANSFORMS
 
 
 # copied from https://github.com/Kitware/nrtk/blob/main/tests/impls/test_pybsm_utils.py
@@ -144,54 +125,3 @@ def create_sample_scenario():
 
 def create_sample_sensor_and_scenario():
     return dict(sensor=create_sample_sensor(), scenario=create_sample_scenario())
-
-
-class NrtkPybsmTransform(ImageTransform):
-    def __init__(self, perturber: PybsmPerturberArg = None):
-        if perturber is None:
-            kwargs = create_sample_sensor_and_scenario()
-            perturber = PybsmPerturber(**kwargs)
-
-        self._perturber: PybsmPerturber = perturber
-
-    def get_parameters(self) -> dict[str, Any]:
-        return {
-            "D": self._perturber.sensor.D,
-            "f": self._perturber.sensor.f,
-        }
-
-    def set_parameters(self, params: Dict[str, Any]):
-        self._perturber.sensor.D = params["D"]
-        self._perturber.sensor.f = params["f"]
-
-    @classmethod
-    def get_parameters_description(cls) -> Dict[str, ParameterDescription]:
-        aperture_description: ParameterDescription = {
-            "type": "float",
-            "label": "Effective Aperture (m)",
-            "default": None,
-            "description": None,
-            "options": None,
-        }
-
-        focal_description: ParameterDescription = {
-            "type": "float",
-            "label": "Focal Length (m)",
-            "default": None,
-            "description": None,
-            "options": None,
-        }
-
-        return {
-            "D": aperture_description,
-            "f": focal_description,
-        }
-
-    def execute(self, input: Image, *input_args: Any) -> Image:
-        if len(input_args) == 0:
-            input_args = ({"img_gsd": 0.15},)
-
-        input_array = np.asarray(input)
-        output_array = self._perturber.perturb(input_array, *input_args)
-
-        return ImageModule.fromarray(output_array)

src/nrtk_explorer/library/nrtk_transforms.yaml

@@ -115,10 +115,10 @@ nrtk_pybsm_detector_otf:
       type: float
       label: Focal length (m)
 
-nrtk_pybsm_2:
+nrtk_pybsm:
   perturber: nrtk.impls.perturb_image.pybsm.perturber.PybsmPerturber
   perturber_kwargs: nrtk_explorer.library.nrtk_transforms.create_sample_sensor_and_scenario
-  exec_default_args: [{ img_gsd: 0.15 }]
+  exec_default_args: [None, { img_gsd: 0.15 }]
   description:
     D:
       _path: [sensor, D]

src/nrtk_explorer/library/yaml_transforms.py

@@ -152,6 +152,6 @@ def execute(self, input, *input_args):
             input_args = self.exec_args
 
         input_array = np.asarray(input)
-        output_array = self._perturber.perturb(input_array, *input_args)
+        output_array, _ = self._perturber.perturb(input_array, *input_args)
 
         return ImageModule.fromarray(output_array)

tests/test_dataset.py

@@ -1,4 +1,4 @@
-from nrtk_explorer.library.dataset import get_dataset, JsonDataset
+from nrtk_explorer.library.dataset import get_dataset, CocoDataset
 import nrtk_explorer.test_data
 
 from pathlib import Path
@@ -27,7 +27,7 @@ def test_get_dataset_empty():
 
 
 def test_DefaultDataset(dataset_path):
-    ds = JsonDataset(dataset_path)
+    ds = CocoDataset(dataset_path)
     assert len(ds.imgs) > 0
     assert len(ds.cats) > 0
     assert len(ds.anns) > 0

tests/test_transforms.py

@@ -14,6 +14,6 @@ def test_gaussian_blur():
 
 def test_pybsm():
     transforms = generate_transforms()
-    pybsm = transforms["nrtk_pybsm_2"]()
+    pybsm = transforms["nrtk_pybsm"]()
     pybsm.set_parameters({"D": 0.25, "f": 4.0})
     pybsm.execute(get_image())