edoardo/modeconverter
1
0
Fork 0
mirror of https://github.com/edoverg/modeconverter.git synced 2026-10-14 23:06:39 +00:00
Code Issues Packages Projects Releases Wiki Activity

added timestamps for evaluation

Browse Source
This commit is contained in:
edoardo
2026-06-17 10:43:26 +02:00
parent e5ae6f6cf2
commit 93be0f6954
1 changed files with 11 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
metaprop_fibermode_doublet.py
+11 -5
View File
@@ -2,11 +2,10 @@
#from jax.numpy.fft import fft2, ifft2, fftshift, ifftshift #from jax.numpy.fft import fft2, ifft2, fftshift, ifftshift
from typing import List, Tuple, Union from typing import List, Tuple, Union
from autograd import numpy as npa, tensor_jacobian_product, grad as gd from autograd import numpy as npa, tensor_jacobian_product, grad as gd
import meep.adjoint as mpa import time
import numpy as np import numpy as np
import matplotlib.pyplot as plt import matplotlib.pyplot as plt
import xarray as xr
from PIL import Image
import nlopt import nlopt
from scipy.special import jv, kv from scipy.special import jv, kv
ArrayLikeType = Union[List, Tuple, np.ndarray] ArrayLikeType = Union[List, Tuple, np.ndarray]
@@ -24,6 +23,7 @@ ds = 1 / res_x
Nx = int(round(res_x * size_x)) + 1 Nx = int(round(res_x * size_x)) + 1
Ny = int(round(res_y * size_y)) + 1 Ny = int(round(res_y * size_y)) + 1
print("Nx: ", Nx)
S = Nx * Ny S = Nx * Ny
norm_phase_min = 0 norm_phase_min = 0
@@ -149,6 +149,7 @@ def forward_propagate(phase_mask:np.ndarray) -> Tuple[List[np.ndarray], List[np.
The output field after propagation (flattened) and the propagation matrices P used in the forward propagation. The output field after propagation (flattened) and the propagation matrices P used in the forward propagation.
''' '''
print("Starting propagating...") print("Starting propagating...")
start_time = time.time()
phase_mask_1 = phase_mask[:S] phase_mask_1 = phase_mask[:S]
phase_mask_2 = phase_mask[S:] phase_mask_2 = phase_mask[S:]
d1 = d[0] d1 = d[0]
@@ -180,7 +181,8 @@ def forward_propagate(phase_mask:np.ndarray) -> Tuple[List[np.ndarray], List[np.
output_field_2_2d = np.fft.ifft2(np.fft.ifftshift(fft_field_propagated_2)) output_field_2_2d = np.fft.ifft2(np.fft.ifftshift(fft_field_propagated_2))
output_field_2 = output_field_2_2d.flatten() output_field_2 = output_field_2_2d.flatten()
print("Propagation finished.") end_time = time.time()
print("Propagation finished in {:.6f} seconds.".format(end_time - start_time))
output_fields = [output_field_1, output_field_2] output_fields = [output_field_1, output_field_2]
propagation_matrices = [P_1, P_2] propagation_matrices = [P_1, P_2]
@@ -196,6 +198,8 @@ def adjoint_propagate(output_fields,propagation_matrices,phase_mask):
Returns: Returns:
The backpropagated field (flattened) The backpropagated field (flattened)
''' '''
print("Computing adjoint...")
start_time = time.time()
phase_mask_1 = phase_mask[:S] phase_mask_1 = phase_mask[:S]
phase_mask_2 = phase_mask[S:] phase_mask_2 = phase_mask[S:]
@@ -232,7 +236,9 @@ def adjoint_propagate(output_fields,propagation_matrices,phase_mask):
grad_C_phi_2 = 2 * np.real(-1j * output_field_1.reshape((Nx,Ny)).T.conj().flatten() * adjoint_field_2_propagated) grad_C_phi_2 = 2 * np.real(-1j * output_field_1.reshape((Nx,Ny)).T.conj().flatten() * adjoint_field_2_propagated)
grads = np.concatenate((grad_C_phi_1, grad_C_phi_2)) grads = np.concatenate((grad_C_phi_1, grad_C_phi_2))
end_time = time.time()
print("Adjoint finished in {:.6f} seconds.".format(end_time - start_time))
return grads return grads
def _compute_cost(phase_mask): def _compute_cost(phase_mask):