radis.test.lbl.test_base module

Created on Mon May 7 17:34:52 2018

@author: erwan

test_export_populations(plot=True, verbose=True, warnings=True, *args, **kwargs)

Check populations calculated in the nonequilibrium module are exported in Spectrum.

Compare with hardcoded values

Examples

::

sf = SpectrumFactory(export_populations=”rovib”)

test_export_rovib_fractions(plot=True, verbose=True, warnings=True, *args, **kwargs)

Compare rovib fraction (nu_vib, nl_vib, nu_rot, nl_rot) are calculated in the nonequilibrium module

Examples

::

sf = SpectrumFactory(export_lines=True, …) sf.misc.export_rovib_fraction = True # required from 0.9.30

test_get_wavenumber_range(*args, **kwargs)

Test waverange conversions for various inputs

test_input_wunit(plot=True, *args, **kwargs)

Test spectrum default units are properly when giving inputs in wavenumber or wavelength ; and test that intensity units are consistent too

test_linestrength_calculations(*args, **kwargs)

Compare and validate linestrength calculations from radis.lbl.base.linestrength_from_Einstein(), radis.lbl.base.BaseFactory.calc_linestrength_eq(), radis.lbl.base.BaseFactory.calc_linestrength_non_eq(),

test_optically_thick_limit_1iso(verbose=True, plot=True, *args, **kwargs)

Test that we find Planck in the optically thick limit

In particular, this test will fail if :

• linestrength are not properly calculated

• at noneq, linestrength and emission integrals are mixed up

The test should be run for 1 and several isotopes, because different calculations paths are used internally, and this can lead to different errors.

Also, this test is used to run with DEBUG_MODE = True, which will check that isotopes and molecule ids are what we expect in all the groupby() loops that make the production code very fast.

Notes

switched from large band calculation with [HITRAN-2016] to a calculation with the embedded [HITEMP-2010] fragment (shorter range, but no need to download files)

test_optically_thick_limit_2iso(verbose=True, plot=True, *args, **kwargs)

Test that we find Planck in the optically thick limit

In particular, this test will fail if :

• linestrength are not properly calculated

• at noneq, linestrength and emission integrals are mixed up

The test should be run for 1 and several isotopes, because different calculations paths are used internally, and this can lead to different errors.

Also, this test is used to run with DEBUG_MODE = True, which will check that isotopes and molecule ids are what we expect in all the groupby() loops that make the production code very fast.

Notes

switched from large band calculation with [HITRAN-2016] to a calculation with the embedded [HITEMP-2010] fragment (shorter range, but no need to download files)

test_populations_CO2_hamiltonian(plot=True, verbose=True, warnings=True, *args, **kwargs)

Calculate nonequilibrium modes with the CO2 Hamiltonian

..warning:

as we only use a reduced set of the CO2 effective Hamiltonian (< 3000 cm-1),
many levels of the Line Database will not appear in the Levels Database.
We will need to either filter the Line Database beforehand, or run it
a first time and remove all levels not found.

This database is obviously not to be used in a Production code!