linting

examples/madelung.py

@@ -7,13 +7,15 @@
 """
 
 import torch
+
 from meshlode import MeshPotential, System
 
+
 # Define simple example structure having the CsCl structure
-positions = torch.tensor([[0,0,0],[0.5,0.5,0.5]])
-atomic_numbers = torch.tensor([55,17]) # Cs and Cl
+positions = torch.tensor([[0, 0, 0], [0.5, 0.5, 0.5]])
+atomic_numbers = torch.tensor([55, 17])  # Cs and Cl
 cell = torch.eye(3)
-charges = torch.tensor([1.,-1.])
+charges = torch.tensor([1.0, -1.0])
 frame = System(species=atomic_numbers, positions=positions, cell=torch.eye(3))
 
 # Define parameters in MeshLODE
@@ -23,10 +25,12 @@
 interpolation_order = 2
 
 # Compute features
-MP = MeshPotential(atomic_smearing=atomic_smearing,
-                   mesh_spacing=mesh_spacing,
-                   interpolation_order=interpolation_order,
-                   subtract_self=True)
+MP = MeshPotential(
+    atomic_smearing=atomic_smearing,
+    mesh_spacing=mesh_spacing,
+    interpolation_order=interpolation_order,
+    subtract_self=True,
+)
 potentials_mesh = MP.compute(frame)
 
 # The ``potentials'' that have been computed so far are not the actual electrostatic
@@ -55,6 +59,6 @@
 # Compare against reference Madelung constant:
 madelung = 2 * 1.7626 / torch.sqrt(torch.tensor(3))
 energies_ref = -madelung * torch.ones((n_atoms, 1))
-print('Computed energies on each atom = \n', atomic_energies)
-print('Reference Madelung constant = \n', madelung)
-print('Total energy = \n', total_energy)
+print("Computed energies on each atom = \n", atomic_energies)
+print("Reference Madelung constant = \n", madelung)
+print("Total energy = \n", total_energy)

src/meshlode/calculators.py

@@ -9,7 +9,7 @@
 Our calculator API follows the `rascaline <https://luthaf.fr/rascaline>`_ API and coding
 guidelines to promote usability and interoperability with existing workflows.
 """
-from typing import Dict, List, Union, Optional
+from typing import Dict, List, Optional, Union
 
 import torch
 from metatensor.torch import Labels, TensorBlock, TensorMap
@@ -33,7 +33,7 @@ class MeshPotential(torch.nn.Module):
     :param atomic_smearing: Width of the atom-centered gaussian used to create the
         atomic density.
     :param mesh_spacing: Value that determines the umber of Fourier-space grid points
-        that will be used along each axis. If set to ``None``, it will automatically
+        that will be used along each axis. If set to None, it will automatically
         be set to half of ``atomic_smearing``
     :param interpolation_order: Interpolation order for mapping onto the grid, where an
         interpolation order of p corresponds to interpolation by a polynomial of degree

src/meshlode/fourier_convolution.py

@@ -2,9 +2,10 @@
 Fourier Convolution
 ===================
 """
-import torch
 from typing import Optional
 
+import torch
+
 
 class FourierSpaceConvolution:
     """
@@ -58,8 +59,7 @@ def generate_kvectors(self, ns: torch.Tensor) -> torch.Tensor:
         return k_vectors
 
     def kernel_func(
-        self, ksq: torch.Tensor, potential_exponent: int = 1,
-        smearing: float = 0.2
+        self, ksq: torch.Tensor, potential_exponent: int = 1, smearing: float = 0.2
     ) -> torch.Tensor:
         """
         Fourier transform of the Coulomb potential or more general effective 1/r**p