visualize verb should find all its inputs from config. (#713)

* `visualize` verb should find all its inputs from config.

Rather that assuming that the dataset provides its own breadcrumbs
to its input, use the session-wide config to locate that data and
its necessary metadata.

* Insist that "infer" be in the data request
* Use setup_dataset from pytorch_ignite in visualize verb
* Replace .numpy() calls with np.asarray() at interface boundaries in visualize.py
* Add REQUIRED_SPLITS and OPTIONAL_SPLITS to Visualize verb

---------

Co-authored-by: Copilot <198982749+Copilot@users.noreply.github.com>
Co-authored-by: gitosaurus <6794831+gitosaurus@users.noreply.github.com>

Author: 2 people

src/hyrax/verbs/visualize.py

@@ -20,6 +20,13 @@ class Visualize(Verb):
     cli_name = "visualize"
     add_parser_kwargs = {}
 
+    # Dataset groups that the Visualize verb knows about.
+    # REQUIRED_SPLITS must be present in the data request configuration.
+    # OPTIONAL_SPLITS are used when present but do not cause an error if absent.
+    # Note: "umap" results may also be required in the future.
+    REQUIRED_SPLITS = ("infer",)
+    OPTIONAL_SPLITS = ()
+
     @staticmethod
     def setup_parser(parser: ArgumentParser):
         """CLI not implemented for this verb"""
@@ -78,7 +85,8 @@ def run(
         from holoviews.streams import Lasso, Params, RangeXY, SelectionXY, Tap
         from scipy.spatial import KDTree
 
-        from hyrax.data_sets.inference_dataset import InferenceDataSet
+        from hyrax.data_sets.result_factories import load_results_dataset
+        from hyrax.pytorch_ignite import setup_dataset
 
         if self.config["data_set"]["object_id_column_name"]:
             self.object_id_column_name = self.config["data_set"]["object_id_column_name"]
@@ -106,10 +114,25 @@ def run(
             )
 
         # Get the umap data and put it in a kdtree for indexing.
-        self.umap_results = InferenceDataSet(self.config, results_dir=input_dir, verb="umap")
-        logger.info(f"Rendering UMAP from the following directory: {self.umap_results.results_dir}")
+        self.umap_results = load_results_dataset(self.config, results_dir=input_dir, verb="umap")
+
+        # Build a DataProvider from the live config for metadata access.
+        # This avoids implicit coupling between result datasets and their original data sources.
+        datasets = setup_dataset(self.config)
+        if not set(Visualize.REQUIRED_SPLITS).intersection(set(datasets.keys())):
+            required_keys = ", ".join(sorted(Visualize.REQUIRED_SPLITS))
+            available_keys = ", ".join(sorted(datasets.keys())) or "<none>"
+            msg = (
+                f"Visualize requires dataset entries {required_keys} in the data request configuration "
+                f"Available dataset keys: {available_keys}"
+            )
+            raise RuntimeError(msg)
+        # NOTE: this presently depends on only a single required
+        # split, but here is here the data provider logic would be
+        # extended if needed.
+        self.metadata_provider = datasets[Visualize.REQUIRED_SPLITS[0]]
 
-        available_fields = self.umap_results.metadata_fields()
+        available_fields = self.metadata_provider.metadata_fields()
         for field in fields.copy():
             if field not in available_fields:
                 logger.warning(f"Field {field} is unavailable for this dataset")
@@ -142,7 +165,7 @@ def run(
         if self.color_column:
             try:
                 # Check if column exists
-                available_fields = self.umap_results.metadata_fields()
+                available_fields = self.metadata_provider.metadata_fields()
                 if self.color_column not in available_fields:
                     logger.warning(
                         f"Column '{self.color_column}' not found in dataset."
@@ -154,7 +177,7 @@ def run(
                     all_indices = list(range(len(self.umap_results)))
 
                     # Extract metadata for the specified column
-                    metadata = self.umap_results.metadata(all_indices, [self.color_column])
+                    metadata = self.metadata_provider.metadata(all_indices, [self.color_column])
                     self.color_values = metadata[self.color_column]
                     logger.info(f"Successfully loaded color values from column '{self.color_column}'")
                     import numpy as np
@@ -308,7 +331,7 @@ def visible_points(self, x_range: Union[tuple, list], y_range: Union[tuple, list
 
         if np.any(np.isinf([x_range, y_range])):
             # Show all points without filtering
-            points = np.array([point.numpy() for point in self.umap_results])
+            points = np.array([np.asarray(point) for point in self.umap_results])
             point_indices = list(range(len(self.umap_results)))
         else:
             # Use existing filtering logic
@@ -358,7 +381,7 @@ def update_points(self, **kwargs) -> None:
             self.points, self.points_id, self.points_idx = self.poly_select_points(kwargs["geometry"])
         elif self._called_tap(kwargs):
             _, idx = self.tree.query([kwargs["x"], kwargs["y"]])
-            self.points = np.array([self.umap_results[idx].numpy()])
+            self.points = np.array([np.asarray(self.umap_results[idx])])
             self.points_id = np.array([list(self.umap_results.ids())[idx]])
             self.points_idx = np.array([idx])
         elif self._called_box_select(kwargs):
@@ -419,7 +442,7 @@ def poly_select_points(self, geometry) -> tuple[npt.ArrayLike, npt.ArrayLike, np
         # Coarse grain the points within the axis-aligned bounding box of the geometry
         (xmin, xmax, ymin, ymax) = Visualize._bounding_box(geometry)
         point_indexes_coarse = self.box_select_indexes([xmin, xmax], [ymin, ymax])
-        points_coarse = self.umap_results[point_indexes_coarse].numpy()
+        points_coarse = np.asarray(self.umap_results[point_indexes_coarse])
 
         tri = Delaunay(geometry)
         mask = tri.find_simplex(points_coarse) != -1
@@ -456,7 +479,7 @@ def box_select_points(
 
         indexes = self.box_select_indexes(x_range, y_range)
         ids = np.array(list(self.umap_results.ids()))[indexes]
-        points = self.umap_results[indexes].numpy()
+        points = np.asarray(self.umap_results[indexes])
         return points, ids, indexes
 
     def box_select_indexes(self, x_range: Union[tuple, list], y_range: Union[tuple, list]):
@@ -495,7 +518,7 @@ def _inside_box(pt):
             return x > xmin and x < xmax and y > ymin and y < ymax
 
         # Filter for points properly inside the box
-        return [i for i in indexes if _inside_box(self.umap_results[i].numpy())]
+        return [i for i in indexes if _inside_box(np.asarray(self.umap_results[i]))]
 
     def selected_objects(self, **kwargs):
         """
@@ -527,7 +550,7 @@ def _table_from_points(self):
 
         # These are the rest of the columns, pulled from metadata
         try:
-            metadata = self.umap_results.metadata(self.points_idx, self.data_fields)
+            metadata = self.metadata_provider.metadata(self.points_idx, self.data_fields)
         except Exception as e:
             # Leave in this try/catch beause some notebook implementations dont
             # allow us to return an exception to the console.
@@ -551,8 +574,10 @@ def _bounding_box(points):
         return (xmin, xmax, ymin, ymax)
 
     def _even_aspect_bounding_box(self):
+        import numpy as np
+
         # Bring aspect ratio to 1:1 by expanding the smaller axis range
-        (xmin, xmax, ymin, ymax) = Visualize._bounding_box(point.numpy() for point in self.umap_results)
+        (xmin, xmax, ymin, ymax) = Visualize._bounding_box(np.asarray(point) for point in self.umap_results)
 
         x_dim = xmax - xmin
         x_center = (xmax + xmin) / 2.0
@@ -585,7 +610,7 @@ def get_selected_df(self):
 
         df = pd.DataFrame(self.points, columns=["x", "y"])
         df[self.object_id_column_name] = self.points_id
-        meta = self.umap_results.metadata(self.points_idx, self.data_fields)
+        meta = self.metadata_provider.metadata(self.points_idx, self.data_fields)
         meta_df = pd.DataFrame(meta, columns=self.data_fields)
 
         cols = [self.object_id_column_name, "x", "y"] + self.data_fields
@@ -656,7 +681,7 @@ def crop_center(arr: np.ndarray, crop_shape: tuple[int, int]) -> np.ndarray:
             sampled_ids = [id_map[idx] for idx in chosen_idx]
 
             # Get metadata - this is in the same order as chosen_idx
-            meta = self.umap_results.metadata(
+            meta = self.metadata_provider.metadata(
                 chosen_idx, [self.object_id_column_name, self.filename_column_name]
             )
 
@@ -696,12 +721,12 @@ def crop_center(arr: np.ndarray, crop_shape: tuple[int, int]) -> np.ndarray:
                         if len(self.torch_tensor_bands) == 1:
                             # Single-band extraction
                             band_idx = self.torch_tensor_bands[0]
-                            arr = tensor[band_idx].numpy()
+                            arr = np.asarray(tensor[band_idx])
                         else:
                             # RGB extraction (3 bands)
                             rgb_arrays = []
                             for band_idx in self.torch_tensor_bands:
-                                rgb_arrays.append(tensor[band_idx].numpy())
+                                rgb_arrays.append(np.asarray(tensor[band_idx]))
                             # Stack along new axis to create (H, W, 3) RGB array
                             arr = np.stack(rgb_arrays, axis=-1)
                     else: