Fetch hitran update

[Supriya1702]

This PR addresses-

• Replacing hit2df with LOCAL_TABLE_CACHE
• added a parameter extra_params in fetch_hitran() and dbmanager.py
• downloading all columns when extra_params = 'all'
• Adding conditions to raise AccuracyWarning when All columns are not downloaded currently.
• cleaning of NaN value columns after downloading all columns in hdf5.py

[erwanp]

Many suggestions above.

Also tests are failing in all nonequilibrium calculations. You can test locally by computing a Spectrum with Tvib different from Trot

[erwanp]

minor comments

[codecov-commenter]

Codecov Report

Merging #505 (fca008e) into develop (b853a26) will increase coverage by 0.19%.
The diff coverage is 69.13%.

@@             Coverage Diff             @@
##           develop     #505      +/-   ##
===========================================
+ Coverage    73.10%   73.30%   +0.19%     
===========================================
  Files          137      138       +1     
  Lines        18870    19015     +145     
===========================================
+ Hits         13795    13939     +144     
- Misses        5075     5076       +1     

[Supriya1702]

@erwanp I was thinking of adding test cases to check if we are able to download the new parameters or not. But am not sure how exactly to implement them, because all the current tests that are being run must be downloading the original database, we can add a test by regenerating the database with extra_params = 'all' or how would you like to proceed with it?

[erwanp]

@erwanp I was thinking of adding test cases to check if we are able to download the new parameters or not. But am not sure how exactly to implement them, because all the current tests that are being run must be downloading the original database, we can add a test by regenerating the database with extra_params = 'all' or how would you like to proceed with it?

Talking about test cases, i suggested #505 (comment)

Good idea to add the new test cases. We'll need them anyway. We shouldn't load a large database, because it will be downloaded on each CI run. But you can try with CO. Does CO have broadneing parameters by other species ? (CO2, etc. ? )

I'd add a fetch_hitran() test in test_hitran_cdsd similarly to how we already test fetch_hitemp() (using "regen") in test_hitemp.py

[anandxkumar]

Great work @Supriya1702, tested it locally, works fine on my end. The only thing missing will be adding a test to check if the extra_params are loading the extra columns or not. I think CO will be a good example of it as it contains gamma_* and n_* for co2, h2o, he (can be used in future for calculating non-air broadening coeff). Maybe we can set extra_params=True and db_use_cached='regen' to check the same in fetch_databank and then test whether columns are loading for load_columns = equilibrium, non-eq and all?

[erwanp]

LGTM

[erwanp]

Well done ; I let you merge when the pipeline succeeds!