MNT: minor linting

.github/workflows/python_package.yml

@@ -34,11 +34,11 @@ jobs:
         # stop the build if there are Python syntax errors or undefined names
         flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics --exclude=src
         # exit-zero treats all errors as warnings. The GitHub editor is 127 chars wide
-        flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics --ignore=E231,E226,E302,E301,E225 --exclude=src
+        flake8 . --count --exit-zero --max-complexity=10 --max-line-length=127 --statistics --ignore=E501,E252,E226,E741,E402,E231,E302,W504,E303,E123,W391,E306,E305
     - name: Lint with Ruff
       run: |
         pip install ruff
-        ruff check --output-format=github --exclude=src .
+        ruff check --output-format=github --ignore E741,E402 .
       continue-on-error: true
     - name: Test with pytest
       run: |

diffeopt/group/ddmatch/action/density.py

@@ -39,7 +39,7 @@ def forward(ctx: Any, x: torch.Tensor, g: Union[torch.Tensor, ForwardDiffeo]) ->
             gy_y[0,:] += shape[1]/2
             gy_y[-1,:] += shape[1]/2
 
-            jac = gx_x*gy_y - gx_y*gy_x # Jacobian
+            jac = gx_x*gy_y - gx_y*gy_x  # Jacobian
 
             pb = np.zeros(np.shape(x))
             compute_pullback(x.detach().numpy(), g_[0], g_[1], pb)

diffeopt/group/ddmatch/action/function.py

@@ -2,7 +2,7 @@
 import torch
 import numpy as np
 
-from ...base import Diffeo, ForwardDiffeo
+from ...base import Diffeo
 
 from ddmatch.core import generate_optimized_image_gradient, generate_optimized_image_composition  # type: ignore
 
@@ -63,7 +63,7 @@ def backward(ctx: Any, grad_output: torch.Tensor) -> tuple[torch.Tensor, Optiona
                 q_ = q.detach().numpy()
                 gxout, gyout = np.zeros_like(q_), np.zeros_like(q_)
                 image_gradient(q_, gxout, gyout)
-                grad = torch.tensor([gxout, gyout])
+                grad = torch.tensor(np.array([gxout, gyout]))
                 result = grad*grad_output
                 return (grad_output, result)
             else:

diffeopt/optim.py

@@ -1,4 +1,4 @@
-from typing import Callable, Any, Optional, Union
+from typing import Callable, Any, Optional
 from abc import ABC, abstractmethod
 import torch
 from .group.representation import Perturbation

diffeopt/utils.py

@@ -22,7 +22,6 @@ def get_random_diffeo(group: BaseDiffeoGroup, nb_steps:int=10, scale:float=1., g
     cometric = laplace.get_laplace_cometric(group, s=2)
     rm = torch.randn(*group.zero().shape, generator=generator)
     rv = cometric(rm)
-    vmx = rv.abs().max()
     shape_scale = (group.shape[0] + group.shape[1])/2
     rv_ = rv/nb_steps/shape_scale**2*scale*32
     current = group.identity()

tests/test_distance.py

@@ -6,7 +6,7 @@ def test_zero_dist():
     """
     Information distance from image to itself is zero.
     """
-    I = 1 + torch.rand(8,8) # make sure it is positive
+    I = 1 + torch.rand(8,8)  # make sure it is positive
     assert information_distance(I,I) == 0
 
 def test_information_derivative():

tests/test_grad.py

@@ -1,4 +1,3 @@
-import numpy as np
 import torch
 from diffeopt.group.ddmatch.action.function import get_composition_action
 from diffeopt.group.ddmatch.action.density import get_density_action
@@ -72,7 +71,6 @@ def test_one_jacobian():
     shape = [16]*2
     group = DiffeoGroup(shape)
     act = get_density_action(shape)
-    idall = group.identity()
     vel = group.zero()
     vel[0] += 3
     trans = group.exponential(vel)