Meow plugin

[simbilod]

Can call MEOW directly on component + layerstack to get single-wavelength Sparameters from EME

• Converts gdsfactory LayerStack to MEOW extrusion rules
• Converts gdsfactory materials to MEOW material
• Converts MEOW output S-parameter format to gdsfactory S-parameter format
• Validates component, currently can only simulate 2-ports components with west and east-facing ports
• Documentation notebook showing simple calls

[sourcery-ai]

Sourcery Code Quality Report

❌  Merging this PR will decrease code quality in the affected files by 0.06%.

Quality metrics	Before	After	Change
Complexity	3.66 ⭐	3.64 ⭐	-0.02 👍
Method Length	116.67 🙂	117.08 🙂	0.41 👎
Working memory	10.41 😞	10.41 😞	0.00
Quality	58.39% 🙂	58.33% 🙂	-0.06% 👎

Other metrics	Before	After	Change
Lines	506	508	2

Changed files	Quality Before	Quality After	Quality Change
gdsfactory/tech.py	58.39% 🙂	58.33% 🙂	-0.06% 👎

Here are some functions in these files that still need a tune-up:

File	Function	Complexity	Length	Working Memory	Quality	Recommendation
gdsfactory/tech.py	LayerStack.get_klayout_3d_script	19 😞	203 😞	22 ⛔	27.65% 😞	Refactor to reduce nesting. Try splitting into smaller methods. Extract out complex expressions
gdsfactory/tech.py	get_layer_stack_generic	0 ⭐	481 ⛔	31 ⛔	34.68% 😞	Try splitting into smaller methods. Extract out complex expressions

Legend and Explanation

The emojis denote the absolute quality of the code:

• ⭐ excellent
• 🙂 good
• 😞 poor
• ⛔ very poor

The 👍 and 👎 indicate whether the quality has improved or gotten worse with this pull request.

---

Please see our documentation here for details on how these metrics are calculated.

We are actively working on this report - lots more documentation and extra metrics to come!

Help us improve this quality report!

[codecov]

Codecov Report

Merging #1015 (2f82ea4) into main (9c77835) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #1015   +/-   ##
=======================================
  Coverage   74.37%   74.38%           
=======================================
  Files         436      436           
  Lines       22521    22522    +1     
  Branches     2987     2987           
=======================================
+ Hits        16751    16752    +1     
  Misses       4878     4878           
  Partials      892      892           

Impacted Files	Coverage Δ
gdsfactory/tech.py	87.58% <100.00%> (+0.08%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[simbilod]

I added a new meow requirement tag to pip install meow, this should be added to docs too

[joamatab]

Thank you Simon!

[joamatab]

fixes #747

[tvt173]

nice!
@flaport

[flaport]

yay, nice work @simbilod .

One comment: in the future we could probably skip meow extrusion conversion and directly define eme cells (following the meow eme cell interface) within GDSFactory. Maybe I'll make an ABC for the Cell interface within meow where people can inherit from.

This would allow for way more general extrusion rules based on the GDSFactory layerstack and possibly tcad cross-section index calculations independent of the meow extrusion rules (which basically are a minimal implementation).

[simbilod]

We could use the existing gmsh plugin to read profiles from arbitrary component cross-sections and LayerStack (see notebooks).

We then could take the mesh (or shapes before meshing) and project them back onto a finite-difference materials grid to use tidy3d to get the EME cells.

Alternatively @HelgeGehring and I made some progress on integrating a FEM mode solver with your EME, but there were some kinks to work out. FEM mode solving is interesting due to adaptive meshing capabilities.

[joamatab]

I think here we should use self.modes = [None] * num_modes

[simbilod]

self.modes contains all of the modes for the EME calculation There is num_modes at each cell. Each list entry in self.modes is another list with num_modes entries I was staying close to the MEOW example to start, we can rename or optimize the silver more. How about self.modes_at_cells?

________________________________ From: Joaquin Matres ***@***.***> Sent: Wednesday, December 21, 2022 8:00:19 AM To: gdsfactory/gdsfactory ***@***.***> Cc: Simon Bilodeau ***@***.***>; Mention ***@***.***> Subject: Re: [gdsfactory/gdsfactory] Meow plugin (PR #1015) @joamatab commented on this pull request.

________________________________ In gdsfactory/simulation/eme/meow_simulation.py<#1015 (comment)>:

+ self.horiz_span = np.diff(component.bbox[:, 1])[0] + horiz_buffer

+ zs = list_unique_layerstack_z(layerstack) + self.vert_span = np.max(zs) - np.min(zs) + vert_buffer + self.horiz_offset = horiz_offset + self.horiz_res = horiz_res + self.vert_offset = vert_offset + self.vert_res = vert_res + + # Setup simulation + self.component, self.layerstack = self.add_global_layers(component, layerstack) + self.extrusion_rules = self.layerstack_to_extrusion() + self.structs = mw.extrude_gds(self.component, self.extrusion_rules) + self.cells = self.get_eme_cells() + self.env = mw.Environment(wl=self.wavelength, T=self.temperature) + self.css = [mw.CrossSection(cell=cell, env=self.env) for cell in self.cells] + self.modes = [None] * num_cells I think here we should use self.modes = [None] * num_modes — Reply to this email directly, view it on GitHub<#1015 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ALCG3WKU2ZD6VHIPQHCP2DDWOMSRHANCNFSM6AAAAAATC63OYI>. You are receiving this because you were mentioned.Message ID: ***@***.***>

[joamatab]

How about self.modes_per_cell?

[simbilod]

Sure, I made it a list so we could only compute cross-section modes as we needed them (for instance, to inspect one before simulating all of them)