vald broadening implementation #225
Conversation
| @@ -21,6 +23,10 @@ | |||
| VACUUM_ELECTRIC_PERMITTIVITY = 1.0 / (4.0 * PI) | |||
There was a problem hiding this comment.
Possibly this is not "vacuum electric permittivity". This looks like a constant not containing epsilon_0, mu_0, or c.
There was a problem hiding this comment.
I agree that this is weird and I can look into it, but it's not really part of the PR and is leftover from something I can't take credit for.
stardis/radiation_field/opacities/opacities_solvers/broadening.py
Outdated
Show resolved
Hide resolved
stardis/radiation_field/opacities/opacities_solvers/broadening.py
Outdated
Show resolved
Hide resolved
stardis/radiation_field/opacities/opacities_solvers/broadening.py
Outdated
Show resolved
Hide resolved
| @@ -414,8 +414,8 @@ def calculate( | |||
| linelist["level_energy_upper"] = ((linelist["e_up"].values * u.eV).cgs).value | |||
|
|
|||
| # Radiation broadening parameter is approximated as the einstein A coefficient. Vald parameters are in log scale. | |||
| linelist["A_ul"] = 10 ** (linelist["rad"]) / ( | |||
| 4 * np.pi | |||
| linelist["A_ul"] = 10 ** ( | |||
There was a problem hiding this comment.
There are 4 places calculating this coefficient in this PR, maybe it's not a problem but if there is a way to optimize it would be nice -- make it a properties of the linelist itself?
There was a problem hiding this comment.
I don't quite know what you're suggesting here. Certainly the molecules and the elements need to be separate functions. The solution I think outside of this PR is just to always use the shortlist calculation, since they're functionally equivalent to the shortlist, but the shortlist needs less information.
But anyway, this is really not part of this PR. There's a bit of standardization slipped in, but this PR mostly doesn't touch the line integrated opacity calculations.
There was a problem hiding this comment.
Oh I see, so the linelist here is already a sliced shortlist of the original list? Sorry I was imaging the same linelist dataset and calculating the A_ul coff multiple times throughout different function. All good then.
This PR lets us use vald broadening parameters and interpolate them to the plasma, rather than calculate them ourselves. On top of the more direct comparison this lets us make to other codes, this also allows us to handle auto-ionization lines (see the line around 5172.4 angstroms below).