Merge pull request #259 from PyAutoLabs/feature/lazy-imports

perf: defer scipy imports to reduce import time

Author: Jammy2211

autoarray/dataset/plot/imaging_plots.py

@@ -1,8 +1,7 @@
 from typing import Optional
 
-import matplotlib.pyplot as plt
 
-from autoarray.plot.utils import subplot_save, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, subplot_save, conf_subplot_figsize, tight_layout
 
 
 def subplot_imaging_dataset(
@@ -51,7 +50,7 @@ def subplot_imaging_dataset(
 
     from autoarray.plot.array import plot_array
 
-    fig, axes = plt.subplots(3, 3, figsize=conf_subplot_figsize(3, 3))
+    fig, axes = subplots(3, 3, figsize=conf_subplot_figsize(3, 3))
     axes = axes.flatten()
 
     plot_array(
@@ -172,7 +171,7 @@ def subplot_imaging_dataset_list(
     from autoarray.plot.array import plot_array
 
     n = len(dataset_list)
-    fig, axes = plt.subplots(n, 3, figsize=conf_subplot_figsize(n, 3))
+    fig, axes = subplots(n, 3, figsize=conf_subplot_figsize(n, 3))
     if n == 1:
         axes = [axes]
     for i, dataset in enumerate(dataset_list):

autoarray/dataset/plot/interferometer_plots.py

@@ -1,12 +1,11 @@
 import numpy as np
 from typing import Optional
 
-import matplotlib.pyplot as plt
 
 from autoarray.plot.array import plot_array
 from autoarray.plot.grid import plot_grid
 from autoarray.plot.yx import plot_yx
-from autoarray.plot.utils import subplot_save, hide_unused_axes, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, subplot_save, hide_unused_axes, conf_subplot_figsize, tight_layout
 from autoarray.structures.grids.irregular_2d import Grid2DIrregular
 
 
@@ -39,7 +38,7 @@ def subplot_interferometer_dataset(
     use_log10
         Apply log10 normalisation to image panels.
     """
-    fig, axes = plt.subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
+    fig, axes = subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
     axes = axes.flatten()
 
     plot_grid(dataset.data.in_grid, ax=axes[0], title="Visibilities", xlabel="", ylabel="")
@@ -117,7 +116,7 @@ def subplot_interferometer_dirty_images(
     use_log10
         Apply log10 normalisation.
     """
-    fig, axes = plt.subplots(1, 3, figsize=conf_subplot_figsize(1, 3))
+    fig, axes = subplots(1, 3, figsize=conf_subplot_figsize(1, 3))
 
     plot_array(
         dataset.dirty_image,

autoarray/fit/plot/fit_imaging_plots.py

@@ -1,9 +1,8 @@
 from typing import Optional
 
-import matplotlib.pyplot as plt
 
 from autoarray.plot.array import plot_array
-from autoarray.plot.utils import subplot_save, symmetric_vmin_vmax, hide_unused_axes, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, subplot_save, symmetric_vmin_vmax, hide_unused_axes, conf_subplot_figsize, tight_layout
 
 
 def subplot_fit_imaging(
@@ -43,7 +42,7 @@ def subplot_fit_imaging(
     grid, positions, lines
         Optional overlays forwarded to every panel.
     """
-    fig, axes = plt.subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
+    fig, axes = subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
     axes = axes.flatten()
 
     plot_array(

autoarray/fit/plot/fit_interferometer_plots.py

@@ -1,11 +1,10 @@
 import numpy as np
 from typing import Optional
 
-import matplotlib.pyplot as plt
 
 from autoarray.plot.array import plot_array
 from autoarray.plot.yx import plot_yx
-from autoarray.plot.utils import subplot_save, symmetric_vmin_vmax, hide_unused_axes, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, subplot_save, symmetric_vmin_vmax, hide_unused_axes, conf_subplot_figsize, tight_layout
 
 
 def subplot_fit_interferometer(
@@ -40,7 +39,7 @@ def subplot_fit_interferometer(
         Not used here (UV-plane residuals are scatter plots); kept for API
         consistency.
     """
-    fig, axes = plt.subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
+    fig, axes = subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
     axes = axes.flatten()
 
     uv = fit.dataset.uv_distances / 10**3.0
@@ -135,7 +134,7 @@ def subplot_fit_interferometer_dirty_images(
     residuals_symmetric_cmap
         Centre residual colour scale symmetrically around zero.
     """
-    fig, axes = plt.subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
+    fig, axes = subplots(2, 3, figsize=conf_subplot_figsize(2, 3))
     axes = axes.flatten()
 
     plot_array(

autoarray/inversion/inversion/inversion_util.py

@@ -5,7 +5,6 @@
 from autoarray.settings import Settings
 
 from autoarray import exc
-from autoarray.util.fnnls import fnnls_cholesky
 
 
 def curvature_matrix_diag_via_psf_weighted_noise_from(
@@ -281,6 +280,8 @@ def reconstruction_positive_only_from(
 
     try:
 
+        from autoarray.util.fnnls import fnnls_cholesky
+
         return fnnls_cholesky(
             curvature_reg_matrix,
             (data_vector).T,

autoarray/inversion/mesh/interpolator/delaunay.py

@@ -1,6 +1,4 @@
 import numpy as np
-import scipy.spatial
-from scipy.spatial import cKDTree, Delaunay, Voronoi
 
 from autoconf import cached_property
 
@@ -12,6 +10,7 @@
 
 def scipy_delaunay(points_np, query_points_np, areas_factor):
     """Compute Delaunay simplices (simplices_padded) and Voronoi areas in one call."""
+    from scipy.spatial import Delaunay
 
     max_simplices = 2 * points_np.shape[0]
 
@@ -182,6 +181,8 @@ def pix_indexes_for_sub_slim_index_delaunay_from(
     # Case 2: Outside → KDTree NN
     # ---------------------------
     if outside_mask.any():
+        from scipy.spatial import cKDTree
+
         tree = cKDTree(delaunay_points)
         _, idx = tree.query(data_grid[outside_mask], k=1)
         out[outside_mask, 0] = idx.astype(np.int32)
@@ -202,6 +203,7 @@ def scipy_delaunay_matern(points_np, query_points_np):
         typically of shape (Q, 3), where each row gives the indices of the
         Delaunay mesh vertices ("pixels") associated with that query point.
     """
+    from scipy.spatial import Delaunay
 
     max_simplices = 2 * points_np.shape[0]
 

autoarray/inversion/plot/inversion_plots.py

@@ -4,12 +4,11 @@
 from pathlib import Path
 from typing import Optional, Union
 
-import matplotlib.pyplot as plt
 from autoconf import conf
 
 from autoarray.inversion.mappers.abstract import Mapper
 from autoarray.plot.array import plot_array
-from autoarray.plot.utils import numpy_grid, numpy_lines, numpy_positions, subplot_save, hide_unused_axes, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, numpy_grid, numpy_lines, numpy_positions, subplot_save, hide_unused_axes, conf_subplot_figsize, tight_layout
 from autoarray.inversion.plot.mapper_plots import plot_mapper
 from autoarray.structures.arrays.uniform_2d import Array2D
 
@@ -53,7 +52,7 @@ def subplot_of_mapper(
     """
     mapper = inversion.cls_list_from(cls=Mapper)[mapper_index]
 
-    fig, axes = plt.subplots(3, 4, figsize=conf_subplot_figsize(3, 4))
+    fig, axes = subplots(3, 4, figsize=conf_subplot_figsize(3, 4))
     axes = axes.flatten()
 
     # panel 0: data subtracted
@@ -279,7 +278,7 @@ def subplot_mappings(
     )
     mapper.slim_indexes_for_pix_indexes(pix_indexes=pix_indexes)
 
-    fig, axes = plt.subplots(2, 2, figsize=conf_subplot_figsize(2, 2))
+    fig, axes = subplots(2, 2, figsize=conf_subplot_figsize(2, 2))
     axes = axes.flatten()
 
     # panel 0: data subtracted

autoarray/inversion/plot/mapper_plots.py

@@ -1,11 +1,10 @@
 import logging
 from typing import Optional
 
-import matplotlib.pyplot as plt
 
 from autoarray.plot.array import plot_array
 from autoarray.plot.inversion import plot_inversion_reconstruction
-from autoarray.plot.utils import numpy_grid, numpy_lines, subplot_save, conf_subplot_figsize, tight_layout
+from autoarray.plot.utils import subplots, numpy_grid, numpy_lines, subplot_save, conf_subplot_figsize, tight_layout
 
 logger = logging.getLogger(__name__)
 
@@ -114,7 +113,7 @@ def subplot_image_and_mapper(
     lines
         Lines to overlay on both panels.
     """
-    fig, axes = plt.subplots(1, 2, figsize=conf_subplot_figsize(1, 2))
+    fig, axes = subplots(1, 2, figsize=conf_subplot_figsize(1, 2))
 
     plot_array(
         image,

autoarray/mask/mask_2d_util.py

@@ -1,6 +1,5 @@
 import numpy as np
 import warnings
-from scipy.ndimage import binary_dilation
 from typing import Tuple
 
 from autoarray import exc
@@ -465,7 +464,6 @@ def min_false_distance_to_edge(mask: np.ndarray) -> Tuple[int, int]:
 from typing import Tuple
 
 import numpy as np
-from scipy.ndimage import binary_dilation
 
 
 def required_shape_for_kernel(
@@ -608,6 +606,8 @@ def blurring_mask_2d_from(
     )
 
     # Pixels within kernel footprint of any unmasked pixel
+    from scipy.ndimage import binary_dilation
+
     near_unmasked_padded = binary_dilation(unmasked_padded, structure=structure)
     near_unmasked = near_unmasked_padded[
         pad_y : pad_y + mask_2d.shape[0],

autoarray/operators/convolver.py

@@ -7,7 +7,6 @@
 
 import numpy as np
 from pathlib import Path
-import scipy
 from typing import Optional, Tuple, Union
 import warnings
 
@@ -116,6 +115,8 @@ class determines how masked real-space data are embedded into a padded array,
         full_shape = tuple(
             s1 + s2 - 1 for s1, s2 in zip(mask_shape, self.kernel.shape_native)
         )
+        import scipy.fft
+
         fft_shape = tuple(scipy.fft.next_fast_len(s, real=True) for s in full_shape)
 
         self.fft_shape = fft_shape