2.1 - 2.2 Migration: ld_coeffs_source

PHOEBE 2.2 introduces the capability to interpolate limb-darkening coefficients for a given ld_func (i.e. linear, quadratic, etc). In order to do so, there is now a new parameter called ld_coeffs_source which will default to 'auto'. The ld_coeffs parameter will not be visibile, unless ld_func is some value other than the default value of 'interp' AND ld_coeffs_source is manually set to 'none'. Any script in which ld_coeffs was set manually, will now require an additional line setting ld_coeffs_source to 'none' (or alternatively removing the line setting ld_coeffs and instead relying on the new capability to interpolate).

Below is an example exhibiting the new behavior.



In [1]:

    
import phoebe
b = phoebe.default_binary()
b.add_dataset('lc', dataset='lc01')
print(b.filter(qualifier='ld*', dataset='lc01'))









    



ParameterSet: 2 parameters
    ld_func@primary@lc01@dataset: interp
  ld_func@secondary@lc01@dataset: interp

By default, ld_func is set to 'interp'. This will interpolate the limb-darkening directly, without requiring a specific law/function.

Note, however, that the bolometric limb-darkening does not have 'interp' as an option. Bolometric limb-darkening is only used for irradiation/reflection, and must be set manually.



In [2]:

    
print(b.filter(qualifier='ld*bol'))









    



ParameterSet: 4 parameters
   ld_func_bol@primary@component: logarithmic
  ld_coeffs_bol@primary@compo...: [0.5 0.5]
  ld_func_bol@secondary@compo...: logarithmic
  ld_coeffs_bol@secondary@com...: [0.5 0.5]

Back to the dataset-specific limb-darkening, we can see the available options besides 'interp'.



In [3]:

    
print(b.get_parameter('ld_func', component='primary').choices)









    



['interp', 'linear', 'logarithmic', 'quadratic', 'square_root', 'power']

And if we set the value of ld_func to anything other than 'interp', we'll now see new parameters for ld_coeffs_source. In PHOEBE 2.1, this would expose the ld_coeffs parameters instead. However, in PHOEBE 2.2+, limb-darkening will be interpolated automatically by default, requiring one extra step to manually set the coefficients.



In [4]:

    
b.set_value_all('ld_func', 'linear')
print(b.filter(qualifier='ld*', dataset='lc01'))









    



ParameterSet: 4 parameters
    ld_func@primary@lc01@dataset: linear
  ld_func@secondary@lc01@dataset: linear
  ld_coeffs_source@primary@lc...: auto
  ld_coeffs_source@secondary@...: auto

Here we see there are several options available for ld_coeffs_source. See the limb-darkening tutorial for more details.



In [5]:

    
print(b.get_parameter('ld_coeffs_source', component='primary').choices)









    



['none', 'auto', 'ck2004']

To manually set the coefficients, we must also set ld_coeffs_source to be 'none'.



In [6]:

    
b.set_value('ld_coeffs_source', component='primary', value='none')
print(b.filter(qualifier='ld*', dataset='lc01'))









    



ParameterSet: 5 parameters
    ld_func@primary@lc01@dataset: linear
  ld_func@secondary@lc01@dataset: linear
  ld_coeffs_source@primary@lc...: none
  ld_coeffs_source@secondary@...: auto
  ld_coeffs@primary@lc01@dataset: [0.5 0.5]

Now that ld_coeffs is visible, run_checks will fail if they are not of the correct length.



In [7]:

    
print(b.run_checks())









    



(False, "ld_coeffs=[0.5 0.5] wrong length for ld_func='linear'.")

By manually setting the value of ld_coeffs to an appropriate value, the checks should pass.



In [8]:

    
b.set_value('ld_coeffs', component='primary', value=[0.5])
print(b.filter(qualifier='ld*', dataset='lc01'))









    



ParameterSet: 5 parameters
    ld_func@primary@lc01@dataset: linear
  ld_func@secondary@lc01@dataset: linear
  ld_coeffs_source@primary@lc...: none
  ld_coeffs_source@secondary@...: auto
  ld_coeffs@primary@lc01@dataset: [0.5]



In [9]:

    
print(b.run_checks())









    



(True, '')