Feature/psf convolution refactor

autogalaxy/__init__.py

@@ -38,6 +38,7 @@
 from autoarray.structures.grids.uniform_2d import Grid2D  # noqa
 from autoarray.structures.grids.irregular_2d import Grid2DIrregular  # noqa
 from autoarray.operators.over_sampling.over_sampler import OverSampler  # noqa
+from autoarray.operators.convolver import Convolver
 from autoarray.inversion.mesh.interpolator.rectangular import (
     InterpolatorRectangular,
 )  # noqa
@@ -48,7 +49,6 @@
 from autoarray.structures.vectors.irregular import VectorYX2DIrregular  # noqa
 from autoarray.layout.region import Region1D  # noqa
 from autoarray.layout.region import Region2D  # noqa
-from autoarray.structures.arrays.kernel_2d import Kernel2D  # noqa
 from autoarray.structures.visibilities import Visibilities  # noqa
 from autoarray.structures.visibilities import VisibilitiesNoiseMap  # noqa
 from autoarray.inversion.mesh.mesh_geometry.rectangular import (

autogalaxy/aggregator/imaging/imaging.py

@@ -59,8 +59,10 @@ def values_from(hdu: int, cls):
         noise_map = values_from(hdu=2, cls=aa.Array2D)
 
         try:
-            psf = values_from(hdu=3, cls=aa.Kernel2D)
+            kernel = values_from(hdu=3, cls=aa.Array2D)
+            psf = aa.Convolver(kernel=kernel)
         except (TypeError, IndexError):
+            kernel = None
             psf = None
 
         dataset = aa.Imaging(

autogalaxy/galaxy/galaxy.py

@@ -455,7 +455,7 @@ def set_snr_of_snr_light_profiles(
         grid: aa.type.Grid2DLike,
         exposure_time: float,
         background_sky_level: float = 0.0,
-        psf: Optional[aa.Kernel2D] = None,
+        psf: Optional[aa.Convolver] = None,
     ):
         """
         Iterate over every galaxy finding all `LightProfileSNR` light profiles and set their `intensity` values to

autogalaxy/galaxy/to_inversion.py

@@ -90,7 +90,7 @@ def _xp(self):
         return np
 
     @property
-    def psf(self) -> Optional[aa.Kernel2D]:
+    def psf(self) -> Optional[aa.Convolver]:
         """
         Returns the PSF of the imaging dataset, if the inversion is performed on an imaging dataset.
 

autogalaxy/imaging/model/plotter_interface.py

@@ -132,7 +132,7 @@ def should_plot(name):
             image_list = [
                 dataset.data.native_for_fits,
                 dataset.noise_map.native_for_fits,
-                dataset.psf.native_for_fits,
+                dataset.psf.kernel.native_for_fits,
                 dataset.grids.lp.over_sample_size.native_for_fits.astype("float"),
                 dataset.grids.pixelization.over_sample_size.native_for_fits.astype(
                     "float"

autogalaxy/imaging/simulator.py

@@ -48,7 +48,7 @@ def via_galaxies_from(
             )
 
         image = galaxies.padded_image_2d_from(
-            grid=grid, psf_shape_2d=self.psf.shape_native
+            grid=grid, psf_shape_2d=self.psf.kernel.shape_native
         )
 
         over_sample_size = grid.over_sample_size.resized_from(
@@ -58,5 +58,5 @@ def via_galaxies_from(
         dataset = self.via_image_from(image=image, over_sample_size=over_sample_size)
 
         return dataset.trimmed_after_convolution_from(
-            kernel_shape=self.psf.shape_native
+            kernel_shape=self.psf.kernel.shape_native
         )

autogalaxy/operate/image.py

@@ -33,7 +33,7 @@ def _blurred_image_2d_from(
         self,
         image_2d: aa.Array2D,
         blurring_image_2d: aa.Array2D,
-        psf: aa.Kernel2D,
+        psf: aa.Convolver,
         xp=np,
     ) -> aa.Array2D:
 
@@ -46,7 +46,7 @@ def blurred_image_2d_from(
         self,
         grid: aa.Grid2D,
         blurring_grid: aa.Grid2D,
-        psf: aa.Kernel2D = None,
+        psf: aa.Convolver = None,
         xp=np,
     ) -> aa.Array2D:
         """
@@ -138,7 +138,7 @@ def unmasked_blurred_image_2d_from(self, grid, psf):
         psf
             The PSF the light object 2D image is convolved with.
         """
-        padded_grid = grid.padded_grid_from(kernel_shape_native=psf.shape_native)
+        padded_grid = grid.padded_grid_from(kernel_shape_native=psf.kernel.shape_native)
 
         padded_image_2d_not_operated = self.image_2d_from(
             grid=padded_grid, operated_only=False
@@ -216,7 +216,7 @@ def blurred_image_2d_list_from(
         self,
         grid: aa.Grid2D,
         blurring_grid: aa.Grid2D,
-        psf: aa.Kernel2D = None,
+        psf: aa.Convolver,
     ) -> List[aa.Array2D]:
         """
         Evaluate the light object's list of 2D images from a input 2D grid of coordinates and convolve each image with
@@ -267,7 +267,7 @@ def blurred_image_2d_list_from(
         return blurred_image_2d_list
 
     def unmasked_blurred_image_2d_list_from(
-        self, grid: aa.Grid2D, psf: aa.Kernel2D
+        self, grid: aa.Grid2D, psf: aa.Convolver
     ) -> List[aa.Array2D]:
         """
         Evaluate the light object's list of 2D images from a input 2D grid of coordinates and convolve it with a PSF,
@@ -290,7 +290,7 @@ def unmasked_blurred_image_2d_list_from(
         psf
             The PSF the light object 2D image is convolved with.
         """
-        padded_grid = grid.padded_grid_from(kernel_shape_native=psf.shape_native)
+        padded_grid = grid.padded_grid_from(kernel_shape_native=psf.kernel.shape_native)
 
         padded_image_2d_list = self.image_2d_list_from(grid=padded_grid)
 

autogalaxy/profiles/light/linear/abstract.py

@@ -142,7 +142,7 @@ def __init__(
         self,
         grid: aa.type.Grid1D2DLike,
         blurring_grid: aa.type.Grid1D2DLike,
-        psf: Optional[aa.Kernel2D],
+        psf: Optional[aa.Convolver],
         light_profile_list: List[LightProfileLinear],
         regularization=Optional[aa.reg.Regularization],
         settings=aa.Settings(),

autogalaxy/profiles/light/snr/abstract.py

@@ -54,7 +54,7 @@ def set_intensity_from(
         grid: aa.type.Grid2DLike,
         exposure_time: float,
         background_sky_level: float = 0.0,
-        psf: Optional[aa.Kernel2D] = None,
+        psf: Optional[aa.Convolver] = None,
     ):
         """
         Set the `intensity` of the light profile as follows:

test_autogalaxy/imaging/test_simulate_and_fit_imaging.py

@@ -15,7 +15,7 @@ def test__perfect_fit__chi_squared_0():
         over_sample_size=1,
     )
 
-    psf = ag.Kernel2D.from_gaussian(
+    psf = ag.Convolver.from_gaussian(
         shape_native=(3, 3), pixel_scales=0.2, sigma=0.75, normalize=True
     )
 
@@ -86,7 +86,7 @@ def test__perfect_fit__chi_squared_0():
 def test__simulate_imaging_data_and_fit__known_likelihood():
     grid = ag.Grid2D.uniform(shape_native=(31, 31), pixel_scales=0.2)
 
-    psf = ag.Kernel2D.from_gaussian(
+    psf = ag.Convolver.from_gaussian(
         shape_native=(3, 3), pixel_scales=0.2, sigma=0.75, normalize=True
     )
 
@@ -144,7 +144,7 @@ def test__simulate_imaging_data_and_fit__linear_light_profiles_agree_with_standa
         over_sample_size=1,
     )
 
-    psf = ag.Kernel2D.from_gaussian(
+    psf = ag.Convolver.from_gaussian(
         shape_native=(3, 3), pixel_scales=0.2, sigma=0.75, normalize=True
     )
 

test_autogalaxy/imaging/test_simulator.py

@@ -52,7 +52,7 @@ def test__from_fits__all_imaging_data_structures_are_flipped_for_ds9():
 
     assert (dataset.data.native == np.array([[0.0, 0.0], [1.0, 0.0]])).all()
     assert (dataset.noise_map.native == np.array([[1.0, 1.0], [2.0, 1.0]])).all()
-    assert dataset.psf.native == pytest.approx(
+    assert dataset.psf.kernel.native == pytest.approx(
         np.array([[0.0, 0.0, 0.0], [0.0, 0.0, 0.0], [0.5, 0.5, 0.0]]), 1.0e-4
     )
 
@@ -83,7 +83,7 @@ def test__from_fits__all_imaging_data_structures_are_flipped_for_ds9():
 
 
 def test__simulator__via_galaxies_from():
-    psf = ag.Kernel2D.from_gaussian(shape_native=(7, 7), sigma=0.5, pixel_scales=0.05)
+    psf = ag.Convolver.from_gaussian(shape_native=(7, 7), sigma=0.5, pixel_scales=0.05)
 
     grid = ag.Grid2D.uniform(shape_native=(20, 20), pixel_scales=0.05)
 
@@ -112,7 +112,7 @@ def test__simulator__via_galaxies_from():
     assert dataset.data.native[10, 10] == pytest.approx(
         imaging_via_image.data.native[10, 10], 1.0e-4
     )
-    assert dataset.psf == pytest.approx(imaging_via_image.psf, 1.0e-4)
+    assert dataset.psf.kernel == pytest.approx(imaging_via_image.psf.kernel, 1.0e-4)
     assert dataset.noise_map == pytest.approx(imaging_via_image.noise_map, 1.0e-4)
 
 
@@ -137,7 +137,8 @@ def test__simulator__simulate_imaging_from_galaxy__source_galaxy__compare_to_ima
 
     grid = ag.Grid2D.uniform(shape_native=(11, 11), pixel_scales=0.2)
 
-    psf = ag.Kernel2D.no_mask(values=[[1.0]], pixel_scales=0.2)
+    kernel = ag.Array2D.no_mask(values=[[1.0]], pixel_scales=0.2)
+    psf = ag.Convolver(kernel=kernel)
 
     simulator = ag.SimulatorImaging(
         psf=psf,
@@ -157,5 +158,5 @@ def test__simulator__simulate_imaging_from_galaxy__source_galaxy__compare_to_ima
 
     assert dataset.shape_native == (11, 11)
     assert dataset.data.array == pytest.approx(imaging_via_image.data.array, 1.0e-4)
-    assert (dataset.psf == imaging_via_image.psf).all()
+    assert (dataset.psf.kernel == imaging_via_image.psf.kernel).all()
     assert dataset.noise_map == pytest.approx(imaging_via_image.noise_map, 1.0e-4)

test_autogalaxy/operate/test_image.py

@@ -95,11 +95,14 @@ def test__x1_galaxies__padded_image__compare_to_galaxy_images_using_padded_grid_
 
 
 def test__unmasked_blurred_image_2d_from():
-    psf = ag.Kernel2D.no_mask(
+
+    kernel = ag.Array2D.no_mask(
         values=(np.array([[0.0, 3.0, 0.0], [0.0, 1.0, 2.0], [0.0, 0.0, 0.0]])),
         pixel_scales=1.0,
     )
 
+    psf = ag.Convolver(kernel=kernel)
+
     mask = ag.Mask2D(
         mask=[[True, True, True], [True, False, True], [True, True, True]],
         pixel_scales=1.0,
@@ -126,7 +129,7 @@ def test__unmasked_blurred_image_2d_from():
     light_not_operated = ag.lp.Gaussian(intensity=1.0)
     light_operated = ag.lp_operated.Gaussian(intensity=1.0)
 
-    padded_grid = grid.padded_grid_from(kernel_shape_native=psf.shape_native)
+    padded_grid = grid.padded_grid_from(kernel_shape_native=psf.kernel.shape_native)
 
     image_2d_not_operated = light_not_operated.image_2d_from(grid=padded_grid)
 
@@ -138,7 +141,7 @@ def test__unmasked_blurred_image_2d_from():
 
     image_2d_operated = padded_grid.mask.unmasked_blurred_array_from(
         padded_array=image_2d_operated,
-        psf=ag.Kernel2D.no_blur(pixel_scales=1.0),
+        psf=ag.Convolver.no_blur(pixel_scales=1.0),
         image_shape=grid.mask.shape,
     )
 
@@ -239,10 +242,11 @@ def test__blurred_image_2d_list_from(
 
 
 def test__unmasked_blurred_image_2d_list_from():
-    psf = ag.Kernel2D.no_mask(
+    kernel = ag.Array2D.no_mask(
         values=(np.array([[0.0, 3.0, 0.0], [0.0, 1.0, 2.0], [0.0, 0.0, 0.0]])),
         pixel_scales=1.0,
     )
+    psf = ag.Convolver(kernel=kernel)
 
     mask = ag.Mask2D(
         mask=[[True, True, True], [True, False, True], [True, True, True]],
@@ -254,7 +258,7 @@ def test__unmasked_blurred_image_2d_list_from():
     lp_0 = ag.lp.Sersic(intensity=1.0)
     lp_1 = ag.lp.Sersic(intensity=2.0)
 
-    padded_grid = grid.padded_grid_from(kernel_shape_native=psf.shape_native)
+    padded_grid = grid.padded_grid_from(kernel_shape_native=psf.kernel.shape_native)
 
     manual_blurred_image_0 = lp_0.image_2d_from(grid=padded_grid)
     manual_blurred_image_0 = psf.convolved_image_from(

test_autogalaxy/profiles/light/test_snr.py

@@ -21,7 +21,7 @@ def test__signal_to_noise_via_simulator_correct():
 
     assert 8.0 < dataset.signal_to_noise_max < 11.5
 
-    psf = ag.Kernel2D.from_gaussian(
+    psf = ag.Convolver.from_gaussian(
         shape_native=(3, 3), sigma=2.0, pixel_scales=0.2, normalize=True
     )