Skip to content

Complete set of optical models #55

Open

Description

@domna

Have at least all models from Fujiwara implemented.

Activity

  1. domna commented on Jun 29, 2022

    @domna
    MemberAuthor

    Implemented

    • Sellmeier
    • Cauchy
    • Lorentz
    • Drude
    • Harmonic Oscillator Approximation
    • Gaussian
    • Tauc-Lorentz

    Need KKR

    • Cody-Lorentz
    • CP Models
    • Polynomials, Splines and B-Spline

    Checked

    This means three things:

    • The docstring is checked (correct formula and general appearance)
    • The formula is the same as in Fujiwara
    • The formula in the docstring and the actual returned value are the same
    • It contains a note if the dispersion is not summable (all n-based dispersions, currently two Cauchys and ConstantRefractiveIndex)

    List

    • Sellmeier
    • Cauchy
    • Lorentz
    • Drude
    • Harmonic Oscillator Approximation
    • Gaussian
    • Tauc-Lorentz
    • Cody-Lorentz
    • CP Models
    • Polynomials, Splines and B-Spline

    Models present in pyElli but not in Fujiwara

    • Poles
    • Tanguy
  2. domna commented on Jun 30, 2022

    @domna
    MemberAuthor

    Maybe also have a look at Sentech and Woolam docs to see what they have to offer and implement that as well.

  3. cfruhling2 commented on Nov 18, 2026

    @cfruhling2

    Hi domna. Thanks so much for your hard work here. In Fujiwara, there is a negative sign in front of the Drude term. The PyElli docs do not show this sign. Should I set A to a negative value to accomplish this?

  4. domna commented on Nov 19, 2026

    @domna
    MemberAuthor

    Yes, typically you would build something like this $\varepsilon = \varepsilon_\infty - \varepsilon_\text{Drude}$ where $\varepsilon_\text{Drude}$ is the dispersion from pyElli's documentation. But of course you might also use a negative $A$. However, in Fujiwara it's $A = \varepsilon_\infty \cdot \omega_p^2$, so $\varepsilon_\infty$ is nonzero and needs to be included in the full term. So it's best to stay with a positive $A$ and use the composite formula (you can just build this in pyElli by writing epsilon = epsilon_infinity - DrudeEnergy(A=..., gamma=....)). It might be a bit awkward here but we decided for pyEllis models to use this adding/subtraction based mechanism instead of fixing $\varepsilon_\infty$ in the dispersion formulas.

    Hope this helps

  5. cfruhling2 commented on Nov 19, 2026

    @cfruhling2

    Many many thanks and much appreciated! I end up with a type error (below) when I subtract. It does not show up when I add.

    Input:

    #generate Drude term
    AZO_dr = elli.dispersions.DrudeEnergy(A = params['AZO_DruAmp'],gamma=params['AZO_DruGamma'])
    
    #generate high frequency term
    AZO_highFreq = elli.dispersions.EpsilonInf(params['AZO_epsInf'])
    
    #set full dispersion
    AZO_disp = AZO_highFreq - AZO_dr
    
    

    This results in the following error :

    TypeError: unsupported operand type(s) for -: 'EpsilonInf' and 'DrudeEnergy'

    This does not happen when I use addition (+)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions