chore: fix numpy.cast deprecation warning

python/examples/example.py

@@ -14,8 +14,9 @@ def add_example_layers(state):
     a[1, :, :, :] = np.abs(np.sin(4 * (iy + iz))) * 255
     a[2, :, :, :] = np.abs(np.sin(4 * (ix + iz))) * 255
 
-    b = np.cast[np.uint32](
-        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10)
+    b = np.asarray(
+        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10),
+        dtype=np.uint32,
     )
     b = np.pad(b, 1, "constant")
     dimensions = neuroglancer.CoordinateSpace(

python/examples/example_coordinate_arrays.py

@@ -14,8 +14,9 @@ def add_example_layers(state):
     a[1, :, :, :] = np.abs(np.sin(4 * (iy + iz))) * 255
     a[2, :, :, :] = np.abs(np.sin(4 * (ix + iz))) * 255
 
-    b = np.cast[np.uint32](
-        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10)
+    b = np.asarray(
+        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10),
+        dtype=np.uint32,
     )
     b = np.pad(b, 1, "constant")
     dimensions = neuroglancer.CoordinateSpace(

python/examples/example_coordinate_transform.py

@@ -14,8 +14,9 @@
     ix, iy, iz = np.meshgrid(
         *[np.linspace(0, 1, n) for n in [100, 100, 100]], indexing="ij"
     )
-    data = np.cast[np.uint32](
-        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10)
+    data = np.asarray(
+        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10),
+        dtype=np.uint32,
     )
     data = np.pad(data, 1, "constant")
     dimensions = neuroglancer.CoordinateSpace(

python/examples/example_local_volume_coordinate_arrays.py

@@ -14,8 +14,9 @@ def add_example_layers(state):
     a[1, :, :, :] = np.abs(np.sin(4 * (iy + iz))) * 255
     a[2, :, :, :] = np.abs(np.sin(4 * (ix + iz))) * 255
 
-    b = np.cast[np.uint32](
-        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10)
+    b = np.asarray(
+        np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10),
+        dtype=np.uint32,
     )
     b = np.pad(b, 1, "constant")
     dimensions = neuroglancer.CoordinateSpace(

python/examples/example_signed_int.py

@@ -10,8 +10,9 @@ def add_example_layer(state):
         *[np.linspace(0, 1, n) for n in [100, 100, 100]], indexing="ij"
     )
     b = (
-        np.cast[np.int32](
-            np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10)
+        np.asarray(
+            np.floor(np.sqrt((ix - 0.5) ** 2 + (iy - 0.5) ** 2 + (iz - 0.5) ** 2) * 10),
+            dtype=np.int32,
         )
         - 2
     )

python/examples/flood_filling_simulation.py

@@ -169,8 +169,8 @@ def process_pos(pos):
                 enqueue(tuple(new_pos))
 
             dist_transform = scipy.ndimage.morphology.distance_transform_edt(~mask)
-            inf_results[slice_expr] = 1 + np.cast[np.uint8](
-                np.minimum(dist_transform, 5) / 5.0 * 254
+            inf_results[slice_expr] = 1 + np.asarray(
+                np.minimum(dist_transform, 5) / 5.0 * 254, dtype=np.uint8
             )
 
             self.viewer.defer_callback(update_view)

python/examples/interactive_inference.py

@@ -96,8 +96,8 @@ def _do_inference(self, action_state):
         boundary_mask[:-1, :, :] |= gt_data[:-1, :, :] != gt_data[1:, :, :]
         boundary_mask[1:, :, :] |= gt_data[:-1, :, :] != gt_data[1:, :, :]
         dist_transform = scipy.ndimage.morphology.distance_transform_edt(~boundary_mask)
-        self.inf_results[slice_expr] = 1 + np.cast[np.uint8](
-            np.minimum(dist_transform, 5) / 5.0 * 254
+        self.inf_results[slice_expr] = 1 + np.asarray(
+            np.minimum(dist_transform, 5) / 5.0 * 254, dtype=np.uint8
         )
         self.inf_volume.invalidate()
 

python/neuroglancer/downsample.py

@@ -31,7 +31,7 @@ def downsample_with_averaging(array, factor):
         indexing_expr = tuple(np.s_[:s] for s in part.shape)
         temp[indexing_expr] += part
         counts[indexing_expr] += 1
-    return np.cast[array.dtype](temp / counts)
+    return np.asarray(temp / counts, dtype=array.dtype)
 
 
 def downsample_with_striding(array, factor):

python/neuroglancer/local_volume.py

@@ -193,7 +193,9 @@ def get_encoded_subvolume(self, data_format, start, end, scale_key):
             or np.prod(downsample_factor) > self.max_downsampling
         ):
             raise ValueError("Invalid downsampling factor.")
-        downsampled_shape = np.cast[np.int64](np.ceil(self.shape / downsample_factor))
+        downsampled_shape = np.asarray(
+            np.ceil(self.shape / downsample_factor), dtype=np.int64
+        )
         if np.any(end < start) or np.any(start < 0) or np.any(end > downsampled_shape):
             raise ValueError("Out of bounds data request.")
 
@@ -208,7 +210,7 @@ def get_encoded_subvolume(self, data_format, start, end, scale_key):
         )
         subvol = np.array(self.data[indexing_expr], copy=False)
         if subvol.dtype == "float64":
-            subvol = np.cast[np.float32](subvol)
+            subvol = np.asarray(subvol, dtype=np.float32)
 
         if np.any(downsample_factor != 1):
             if self.volume_type == "image":

testdata/generate_npy_examples.py

@@ -36,11 +36,11 @@ def write_array(array):
         np.save(name, new_array)
     array_for_json = array
     if dtype == np.uint64:
-        array_for_json = (np.cast[np.dtype("<u8")](array)).view("<u4")
+        array_for_json = np.asarray(array, dtype=np.dtype("<u8")).view("<u4")
     if dtype == np.float32:
-        array_for_json = np.cast[float](array_for_json)
+        array_for_json = np.asarray(array_for_json, dtype=float)
     else:
-        array_for_json = np.cast[int](array_for_json)
+        array_for_json = np.asarray(array_for_json, dtype=int)
     json_type = float if dtype.kind == "f" else int
     json_obj = dict(
         dataType=dtype.name,
@@ -65,4 +65,4 @@ def write_int_array(dtype):
 write_int_array(np.uint16)
 write_int_array(np.uint32)
 write_int_array(np.uint64)
-write_array(np.cast[np.float32](gen.standard_normal(shape)))
+write_array(np.asarray(gen.standard_normal(shape), dtype=np.float32))