demisjohn / pyfimm Goto Github PK

• Bend Loss calc - auto-polishing with FDM->FMM + PML -> FMM + Transp (Hands-on training p.107, example 2B)

Open WG from File, create new object by copy/pasting the dev. into project, and populate the WG params.

• Waveguide - findorcreateviews() + updateviews()

Import from File - reference path inside a file, import into new object. ImportWG, ImportCirc, ImportDev()

__pyfimm.py line 424: Remove all instances of .strip()[:-2] and use if xyz.endswith() // xyz.strip(etc)

• Check that Circ & Device can take same keyworded args as Waveguide() & Waveguide.buildNode()

- Circ - support Gaussian & Spline index profiles

Instead of using Node.__checkNodeName()

Why does Nmodes affect symmetry of Cavity e-field? Is EigVect captured incorrectly?

• WG.plot_refractive_index - also possible to include in Mode plots via refractive_index=True

• Implement calcevlisterr() - orthogonality test

• Mode check for max modes calc'd (see nevals)

• Update docstrings to numpydoc format

• Mode__len__ returns number of calc'd modes (None for not clac'd)

Open WG from File, create new object by copy/pasting the WG into project

• Lens - make sure Section gets a Section.length attribute for a Lens

• Bend Loss calc - auto-polishing with FDM->FMM + PML -> FMM + Transp (Hands-on training p.107, example 2B)

• Option for MOLAB search start & end n_eff (global pyfimm settings). applicable to FEM, FDM - not sure of others. If only start is spec'd, will search for this OR LESS. Maybe take two-valued tuple as arg? Or separate set_start() & set_end()

• paths/curved waveguides?

From Vincent Brulis: Yes you just need to sum up the field amplitudes. If you want to normalise your vector you simply need to normalise the field coefficients; the normalised vector corresponding to [1.0 + 0j , 0.5 + 0.5j , 0 , 0] is [1.0/sqrt(1.5) + 0j , 0.5/sqrt(1.5) + 0.5j/sqrt(1.5) , 0 , 0]

• Device - set Device-level Solver etc. params? Answer - no, leave it alone

• set_X_pml()- if pass a 2-tuple, set left & right differently. Same for Y_pml()

• mode.plot() - Intensity 'I' plot

Cavity.Calc

• refine=True option, to automatically minimize image(EigV) closest to zero, within some tolerance (and add the additional calc'd wavelengths to the wavelengths attribute)

Open Device from File, create new object by copy/pasting the dev. into project

• Circ- Load arbitrary index profile from file

• Temperature support - in WG node builders & WG-specific temperatures (global get/set already implemented)

• Circ - Make sure all solver types work

-- __pyfimm.Node - should check the correct subnode level for name conflicts. All nodes are now in root-level of the project, so checking sublevels is unnecessary.

• IOports objects in Devices

Make sure all checks of NodeName use one function, pyfimm.checkNodeName().

• allow Calc (WG,Circ,Dev) to take arguments for mode solver (update mlp's on the fly - requires that global MoLab Params have local (WG./Dev.) equivalent first)

• consistency in nodestring having trailing . - may be done already

Device take other Device obj as element

• set_mode_solver() & Circ/WG buildNodes- process mode solver string and break into boolean keys etc.

• Add global params into WG & Device - set_N, set_solver, set_N1d etc…
• All global params should have local (WG,Dev) equiv., so they can be changed before/after building for each element

• Material params (index, LOSS) for

• Device-> Periodic sections. Use New-SubElement & New-SubJoint to set the reference device to repeat & joint between them

(most recent call last):
  File "<stdin>", line 1, in <module>
  File "pyfimm\__pyfimm.py", line 424, in buildNode
    SNnames.append(  fimm.Exec(r"app.subnodes["+str(i+1)+"].nodename").strip()[:-2]  )   
AttributeError: 'float' object has no attribute 'strip'

typecast fimmwave output to string if a name is expected? Should have done:

SNnames.append(  str(  fimm.Exec(r"app.subnodes["+str(i+1)+"].nodename").strip()[:-2]  )  )

• Circ/WG - load WG from File into node, for use in Device. (allows for manual construction in FimmWave GUI)

• Circ - Spline index profile

• Mode returns "None" if specified mode doesn't exist (so Mode(0)=None means not calc'd)

• Device.set_input()- is integer, set input mode number. If None, turn off input

From Vincent Brulis:

We have upgraded the command line interface to allow it to call functions inline, for instance:
app.[…].fdtd.fluxvposition(…).get()
in OmniSim.

You can find more details in the Changes Logs of the new FIMMWAVE manual once you receive it.


Add field component option for plotting field squared, eg. 'Ex2', 'Ey2' etc.etc. Probably just use string check to see if last char is '2', eg. if component.endswith('2'):...

• Add global params into WG & Device - set_N, set_solver, set_N1d etc…
• All global params should have local (WG,Dev) equiv., so they can be changed before/after building for each element

• Reconicle self.length with Section(WG() ) - maybe self.length should go away, or be used when WG is called with no args.

• Calc(polish=True) - automatic polishing.... somehow. Just set the "polish" flag in solver params?
  requires that Calc can take Molab params, so the polish would have an effect (eg. high res, diff. mode solver etc.)

• Dev.set_input_field() - accept Mode number mode (to turn off an input)