fix femwell and try to fix meow plugin

[joamatab]

• fix femwell installation
• eme plugin still has some issues to simulate a taper

[codecov]

Codecov Report

Merging #1038 (9102d1a) into main (b2a4ee7) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #1038   +/-   ##
=======================================
  Coverage   72.57%   72.57%           
=======================================
  Files         436      436           
  Lines       22543    22543           
  Branches     2988     2988           
=======================================
  Hits        16361    16361           
  Misses       5305     5305           
  Partials      877      877           

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

[joamatab]

#1015

[simbilod]

I've noticed that results have a very strong dependence on the simulation parameters (num_modes, resolution, cell_length)

It helps to use negative spacing_y to make the cross_section a bit smaller, and hence require coarser grid, than what the LayerStack would give you by default.

AFAIK basic EME like this needs a lot of slices to avoid spurious reflections due to "staircasing". Commercial EME software usually employ some sort of first-order interpolation scheme to avoid this and converge the results better.

[simbilod]

See e.g.

` c = gf.components.taper(length=10, width2=2)
c.show()

from gdsfactory.tech import get_layer_stack_generic

filtered_layerstack = LayerStack(
    layers={
        k: get_layer_stack_generic().layers[k]
        for k in (
            "slab90",
            "core",
            "box",
            "clad",
        )
    }
)

trans = []
cells_lengths = [0.1,0.25,0.5,0.75,1]
for cell_length in cells_lengths:

    m = MEOW(component=c, 
                layerstack=filtered_layerstack, 
                wavelength=1.55,
                resolution_x=100,
                resolution_y=100,
                spacing_x=1,
                spacing_y=-3,
                num_modes=4,
                cell_length=cell_length,
                )
    import matplotlib.pyplot as plt
    # m.plot_mode(xs_num = -1, mode_num = 0)
    # m.plot_cross_section(xs_num = -1)
    sp = m.compute_sparameters()
    te0_trans = np.abs(sp["o1@0,o2@0"])**2
    trans.append(te0_trans)
    # print(len(m.cells))

plt.plot(cells_lengths, trans)
plt.title("10um taper, resx = resy = 100, num_modes = 4")
plt.xlabel("Cell length (um)")
plt.xlabel("TE0 transmission")
plt.show()`

[simbilod]

x-axis should be cell_size from 100 nm to 1 um, and y axis transmission **

[joamatab]

This is great 😊

I see, how about reduce the default cell length to 0.5?

[simbilod]

Sure!

[sourcery-ai]

Sourcery Code Quality Report

✅  Merging this PR will increase code quality in the affected files by 1.49%.

Quality metrics	Before	After	Change
Complexity	2.44 ⭐	2.19 ⭐	-0.25 👍
Method Length	76.58 🙂	79.30 🙂	2.72 👎
Working memory	11.04 😞	10.44 😞	-0.60 👍
Quality	65.51% 🙂	67.00% 🙂	1.49% 👍

Other metrics	Before	After	Change
Lines	305	372	67

Changed files	Quality Before	Quality After	Quality Change
gdsfactory/simulation/eme/meow_eme.py	66.56% 🙂	67.22% 🙂	0.66% 👍
gdsfactory/simulation/eme/test_meow_simulation.py	58.44% 🙂	66.03% 🙂	7.59% 👍

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

File	Function	Complexity	Length	Working Memory	Quality	Recommendation
gdsfactory/simulation/eme/meow_eme.py	MEOW.__init__	0 ⭐	250 ⛔	22 ⛔	41.92% 😞	Try splitting into smaller methods. Extract out complex expressions
gdsfactory/simulation/eme/test_meow_simulation.py	test_meow_defaults	8 ⭐	139 😞	12 😞	52.64% 🙂	Try splitting into smaller methods. Extract out complex expressions
gdsfactory/simulation/eme/meow_eme.py	MEOW.compute_sparameters	3 ⭐	157 😞	12 😞	55.57% 🙂	Try splitting into smaller methods. Extract out complex expressions
gdsfactory/simulation/eme/meow_eme.py	MEOW.gf_material_to_meow_material	4 ⭐	109 🙂	10 😞	63.81% 🙂	Extract out complex expressions
gdsfactory/simulation/eme/meow_eme.py	MEOW.add_global_layers	3 ⭐	112 🙂	10 😞	64.41% 🙂	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!