search

SunnySuite / Sunny.jl

Spin dynamics and generalization to SU(N) coherent states
Other
86 stars 19 forks source link

Add SU(N) LSWT to Sunny #68

Closed Hao-Phys closed 1 year ago

Hao-Phys commented 1 year ago

This PR adds the SU(N) linear spin-wave theory features to Sunny. The :SUN and : dipole modes are supported. But the :large_S mode is not supported yet. Moreover, the biquad interactions are not supported.

kbarros commented 1 year ago

Hi Hao, thanks for the PR. Overall this looks great to me. My suggestions are mostly concerning the public-facing user interface. Maybe we should discuss together with @ddahlbom . The implementation details can always be revised later, so I want to focus for now one the question of what exported functions we should be exposing.

kbarros commented 1 year ago

Are we good to merge now with some minor renamings? We can rename more later in a follow up.

Hao-Phys commented 1 year ago

What are the renamings we need to do right now?

ddahlbom commented 1 year ago

I'm just thinking about making the API as similar as possible to what its final form will be and making it so that the FeI2 example will look as clean as possible in the meantime. In this spirit, and based on the discussion above, I would suggest:

1) lswt_unpolarized_INS_spec -> intensities. 2) lswt_dispersion_relation -> dispersion. 3) lswt_dynamical_spin_structure_factor! -> dssf. It would also be good if the Sαβ_matrix could be allocated internally. (Another possible name for this function would be dynamic_structure_factor, but I think it's a bit odd that we would briefly have both dynamic_structure_factor and DynamicStructureFactor.) 4) SpinWaveFields -> SpinWave

The naming/interface for (3) and (4) will probably be subject to the most changes later on, but I think these changes would be good for now.

kbarros commented 1 year ago

These suggestions look good to me. Thanks all, let's get it merged 👍

Kipton Barros, from phone

On Sun, Mar 19, 2023, 12:52 David A. Dahlbom @.***> wrote:

I'm just thinking about making the API as similar as possible to what it's final form will be and making it so that the FeI2 example will look as clean as possible. In this spirit, and based on the discussion above, I would suggest:

  1. lswt_unpolarized_INS_spec -> intensities.
  2. lswt_dispersion_relation -> dispersion.
  3. lswt_dynamical_spin_structure_factor! -> dssf. It would also be good if the Sαβ_matrix could be allocated internally. (Another possible name for this function would be dynamic_structure_factor, but I think it's a bit odd that we would briefly have both dynamic_structure_factor and DynamicStructureFactor.)
  4. SpinWaveFields -> SpinWave

The naming/interface for (3) and (4) will probably be subject to the most changes later on, but I think those changes would be good for now.

— Reply to this email directly, view it on GitHub https://github.com/SunnySuite/Sunny.jl/pull/68#issuecomment-1475319227, or unsubscribe https://github.com/notifications/unsubscribe-auth/AACG3HM325ZFEKNIGS5AHSDW442VLANCNFSM6AAAAAAV57FKLQ . You are receiving this because you commented.Message ID: <SunnySuite/Sunny .@.***>

Hao-Phys commented 1 year ago

I have changed the function names based on David's suggestions. Moreover, I have added supports for biquad in the :dipole mode (and a corresponding test item in test_lswt.jl). We are ready to merge.