Utility for writing to IMAS

[eldond]

Hi,
I have https://github.com/ProjectTorreyPines/SD4SOLPS.jl/blob/24af0ba75505bc56ce9a57b122c8cd9fcc721383/src/SD4SOLPS.jl#L78 which uses EFIT.jl output and translates data into the IMAS schema. I developed this within one of the repos from our project, but maybe it makes more sense as part of EFIT.jl? IMASDD is already a dependency. Let me know if I should move this function. Maybe it would go in io.jl?

[orso82]

@eldond I would avoid adding derived quantities to dd, since that's something that the IMAS.jl package will be able to do directly from the data in dd. Also, be careful to always load data in COCOS 11 when it goes into dd. gEQDSKs can have 4 different COCOS conventions: https://docs.google.com/document/d/1-efimTbI55SjxL_yE_GKSmV4GEvdzai7mAj5UYLLUXw/edit

Perhaps this would be a better starting point:

https://github.com/ProjectTorreyPines/FUSE.jl/blob/955016241ca32a17b471f2a604ae90e4ca1c1b9f/src/actors/equilibrium/chease_actor.jl#L158

"""
    gEQDSK2IMAS(GEQDSKFile::GEQDSKFile,eq::IMAS.equilibrium)

Convert IMAS.equilibrium__time_slice to MXHEquilibrium.jl EFIT structure
"""
function gEQDSK2IMAS(g::CHEASE.EFIT.GEQDSKFile, eq::IMAS.equilibrium)
    tc = CoordinateConventions.transform_cocos(1, 11) # chease output is cocos 1 , dd is cocos 11

    eqt = eq.time_slice[]
    eq1d = eqt.profiles_1d
    eq2d = resize!(eqt.profiles_2d, 1)[1]

    @ddtime(eq.vacuum_toroidal_field.b0 = g.bcentr)
    eq.vacuum_toroidal_field.r0 = g.rcentr

    eqt.global_quantities.magnetic_axis.r = g.rmaxis
    eqt.boundary.geometric_axis.r = g.rcentr
    eqt.boundary.geometric_axis.z = g.zmid
    eqt.global_quantities.magnetic_axis.z = g.zmaxis
    eqt.global_quantities.ip = g.current

    eq1d.psi = g.psi .* tc["PSI"]
    eq1d.q = g.qpsi
    eq1d.pressure = g.pres
    eq1d.dpressure_dpsi = g.pprime .* tc["PPRIME"]
    eq1d.f = g.fpol .* tc["F"]
    eq1d.f_df_dpsi = g.ffprim .* tc["F_FPRIME"]

    eq2d.grid_type.index = 1
    eq2d.grid.dim1 = g.r
    eq2d.grid.dim2 = g.z
    eq2d.psi = g.psirz .* tc["PSI"]

    return nothing
end

[eldond]

Okay. Should the writing function be in EFIT or in IMAS, though?

[orso82]

Whether you mean efit2imas or imas2efit I think those should exist in the EFIT.jl package

[eldond]

This stops us from importing EFIT into IMAS later because it would create a circular dependency, right?

[orso82]

@eldond we do not need nor want to import EFIT into IMAS. Is there some specific use case that you have in mind?

[eldond]

@orso82 I want to import data from an eqdsk file into IMAS. I am using EFIT.jl to do this so far.

[orso82]

EFIT.jl with IMASdd.jl as a dependency, and efit2imas and imas2efit functions will be able to do that. No circular dependency needed.