Running cobaya using Planck high-l TE or EE likelihoods

[Minh Nguyen]

Hi Jesus and Antony,

I'm a cobaya novice user and trying to reproduce some of the plots in the Planck2018 cosmo paper.

I noticed that there's pre-defined Planck2018 high-l likelihoods for TT and TT+TE+EE, but there isn't one for high-l TE or EE separately.

All the Planck2018 likelihood definitions I found inside cobaya code seem to be python wrappers of the Planck clik versions for both high-l TT, TE and EE.

How should I go about adding these options for using Planck2018 high-l TE and/or EE separately? Do I define a new likelihood class or is it just about adding those options into the input database?

Many thanks,
Minh

[Antony Lewis]

There should be separate EE and TE files under the planck_2018_highl_plik likelihoods.

[Minh Nguyen]

Thanks, Antony. Right, my question is how can I make use of those files under

Code: Select all

likelihoods/planck_2018_highl_plik/

?

So I can just add the keys

Code: Select all

"planck_2018_highl_plik.TE": None,
"planck_2018_highl_plik.EE": None,

in

Code: Select all

cobaya/cosmo_input/input_database.py

?

Thank you for your patience!

[Antony Lewis]

Probably easier to just edit the output .yaml as required

[Minh Nguyen]

Thank you! I just try to specify those likelihoods in the yaml file, and once I make sure that the extended data is at the right path, cobaya seems to work fine.