incorrect PMLs for Cartesian coordinates in plot2D

[oskooi]

#1873 seems to have introduced a bug in plot2D for PMLs in Cartesian coordinates. The bug appears specifically for PMLs in the meep.X direction with side meep.Low as demonstrated in the following example. In the example, there should be a PML in the -x boundary of the cell. Separately, it would be good to add a unit test for plot2D to avoid these types of bugs in the future.

PR 1873 (missing PML in left boundary)

prior to PR 1873

from meep.materials import Au
import meep as mp
import numpy as np
import matplotlib

n_H2O = 1.333 # refractive index of water                                                                                                                    

r = 1.0  # radius of sphere                                                                                                                                  

wvl_min = 2*np.pi*r/10
wvl_max = 2*np.pi*r/2

frq_min = 1/wvl_max
frq_max = 1/wvl_min
frq_cen = 0.5*(frq_min+frq_max)
dfrq = frq_max-frq_min
nfrq = 100

resolution = 100

dpml = 0.5*wvl_max
dair = 0.5*wvl_max

pml_layers = [mp.PML(thickness=dpml)]

s = 2*(dpml+dair+r)
cell_size = mp.Vector3(s,s,0)

sources = [mp.Source(mp.GaussianSource(frq_cen,fwidth=dfrq,is_integrated=True),
                     center=mp.Vector3(-0.5*s+dpml),
                     size=mp.Vector3(0,s,0),
                     component=mp.Ez)]

symmetries = [mp.Mirror(mp.Y)]

geometry = [mp.Sphere(material=Au,
                      center=mp.Vector3(),
                      radius=r)]

sim = mp.Simulation(resolution=resolution,
                    cell_size=cell_size,
                    boundary_layers=pml_layers,
                    sources=sources,
                    k_point=mp.Vector3(),
                    symmetries=symmetries,
                    default_material=mp.Medium(index=n_H2O),
                    geometry=geometry)

box_x1 = sim.add_flux(frq_cen, dfrq, nfrq, mp.FluxRegion(center=mp.Vector3(x=-r),size=mp.Vector3(0,2*r)))
box_x2 = sim.add_flux(frq_cen, dfrq, nfrq, mp.FluxRegion(center=mp.Vector3(x=+r),size=mp.Vector3(0,2*r)))
box_y1 = sim.add_flux(frq_cen, dfrq, nfrq, mp.FluxRegion(center=mp.Vector3(y=-r),size=mp.Vector3(2*r,0)))
box_y2 = sim.add_flux(frq_cen, dfrq, nfrq, mp.FluxRegion(center=mp.Vector3(y=+r),size=mp.Vector3(2*r,0)))

sim.plot2D()
plt.savefig('bug_plot2D.png',dpi=150,bbox_inches='tight')

cc @smartalecH

[smartalecH]

#1873 seems to have introduced a bug in plot2D for PMLs in Cartesian coordinates.

Whoops. Easy fix.

The lines here:

meep/python/visualization.py

Lines 480 to 481 in d00febd

	if (permutation[0] == mp.X) and (permutation[1] == mp.Low):
	continue

here:

meep/python/visualization.py

Lines 487 to 488 in d00febd

	if (boundary.direction == mp.X) and (side == mp.Low):
	continue

and here:

meep/python/visualization.py

Lines 493 to 494 in d00febd

	if (boundary.direction == mp.X) and (boundary.side == mp.Low):
	continue

just need to check for cylindrical coordinates (like here):

meep/python/visualization.py

Line 471 in d00febd

elif sim.dimensions == mp.CYLINDRICAL or sim.is_cylindrical:

Separately, it would be good to add a unit test for plot2D to avoid these types of bugs in the future.

We already have a unit test (see here). There's just no easy way to test if an image is "right".

We currently hash the image and make sure it's reasonably similar to what's new.