broadening parameters seem to only use J" quantum number and ignore other quantum numbers in the "*.broad" files.

[jackdobinson]

I have been exploring how to get line data out of PyExoCross, and I may have found a slight bug. The broadening parameters seem to only be matched w.r.t the J" quantum number, not any of the other quantum numbers defined in the QN-set of each "*.broad" file. I came across this while trying to read a line list out in HITRAN format, but get broadening parameters for not just "self" and "air".

Details

Looking in the file ".../PyExoCross/src/pyexocross/database/load_exomol.py", the two functions read_broad(...) and extract_broad(...) seem to be the only places where data from the "*.broad" files are handled, but the dataframes produced by read_broad(...) only have a "Jpp" column and no other quantum numbers. And extract_broad(...) only operates on the "Jpp" quantum number column instead of matching all the QNs in a set.

[natashabatalha]

Echoing this comment. Adding more details..

For example consider 14N-1H3 broadened by H2 and He. The .broad file in Table 22 the exocross paper only considers 4 columns to Jpp. But the same file on the exocross website now contains more columns. Therefore, this line of code returns an error:

https://github.com/Beryl-Jingxin/PyExoCross/blob/9529aa04c4a9e65a91aed2d678371e341f167d69/src/pyexocross/database/load_exomol.py#L424

broad_df = pd.read_csv(fname_broadener, sep='\s+', header=None, engine='python')

changing to

broad_df = pd.read_csv(fname_broadener, sep=rf'\s+', header=None, engine='python',usecols=[0,1,2,3])

seems to pass but I wanted to confirm this is intentional.

[Beryl-Jingxin]

Sorry that I missed the issues in this repository.

Thank you both for reporting this issue. It's caused by the ExoMol2024 new broadening file format. I have updated the PyExoCross to support extra columns in .broad files (usually totally have 4 or 5 columns).

If there is still have any problem, please let me know.

An example for NH3:

import pyexocross as px

COMMON = dict(
    database='ExoMol',
    molecule='NH3',
    isotopologue='14N-1H3',
    dataset='CoYuTe',
    species_id=111,
    read_path='/Users/beryl/Academic/UCL/PhD/Data/database/ExoMol/', 
    save_path='/Users/beryl/Academic/UCL/PhD/Data/pyexocross/', 
    logs_path='/Users/beryl/Academic/UCL/PhD/Data/pyexocross/log/NH3_ExoMol_broad.log', 
)


# Spectral range parameters
RANGE_PARAMS = dict(
    temperatures=[1175.0],              # Temperature in unit of K
    wn_wl='WN',                         # Wavenumber (wn in unit cm⁻¹) or wavelength (wl in unit[nm or um]) (default: 'WN')
    wn_wl_unit='cm-1',                  # Unit for wavenumber (default: cm⁻¹)
    min_range=3330.0,                   # Minimum wavenumber in unit of cm⁻¹
    max_range=3339.99695288349,         # Maximum wavenumber in unit of cm⁻¹
    abs_emi='Ab',                       # Absorption or emission (default: 'Absorption')
    unc_filter=None,                    # Uncertainty filter (default: None)
    threshold=1e-30,                    # Threshold filter (default: None)
)

# Cores and chunks
COMPUTE_PARAMS = dict(
    ncputrans=4,                # Number of CPU threads for each transition file (default: 4)
    ncpufiles=1,                # Number of CPU files for transition calculation (default: 1)
    chunk_size=500000,          # Chunk size for transition calculation (default: 100000)
    run_mode='CPU',             # Run mode: 'CPU' or 'GPU' (default: 'CPU')
)

px.cross_sections(
      **COMMON,
      **RANGE_PARAMS,
      **COMPUTE_PARAMS,
      pressures=[1e-06],      # Pressure in unit bar (default: [1.0])
      n_point=2854,           # Bin size for wavenumber grid 
      profile='SciPyVoigt',   # Line profile name (default: 'Gaussian')
      predissociation=False,  # Predissociation (default: False)
      cutoff=30,              # Cutoff distance in cm⁻¹ (default: None)
      broadeners=['H2', 'He'],# Broadening species (default: ['Default'])
      ratios=[0.85, 0.15],    # Broadening ratios (default: [1.0])
      plot=True,              # Whether to plot results (default: False)
      plot_method='log',      # Plot in linear (lin) or logarithm (log) (default: 'log')
      plot_wn_wl='WN',        # Wavenumber (wn in unit cm⁻¹) or wavelength (wl in unit[nm or um]) (default: 'WN')
      plot_unit='cm-1',       # Unit for plotting axis (default: cm⁻¹)
      limit_yaxis=1e-21,      # Lower limit for y-axis (default: 1e-30 cm²/molecule)
    )

[Beryl-Jingxin]

Please update to version 1.1.6