ACT DR6 Multifrequency Likelihood

An external likelihood using cobaya and based on LAT_MFLike.

Installing the code

If you do not plan to dig into the code, you can simply install the code with the following command

python -m pip install act_dr6_mflike [--user]

Otherwise, you first need to clone this repository to some location

git clone https://github.com/ACTCollaboration/act_dr6_mflike.git /where/to/clone

Then you can install the act_dr6_mflike likelihood and its dependencies via

python -m pip install -e /where/to/clone

Installing ACT DR6 data

To install ACT DR6 data, the easiest way is to use the cobaya-install binary and let it do the installation job. For instance, if you do the next command

cobaya-install /where/to/clone/examples/act_dr6_example.yaml -p /where/to/put/packages

data and code such as CAMB will be downloaded and installed within the /where/to/put/packages directory. For more details, you can have a look to cobaya documentation.

Running/testing the code

You can test the act_dr6_mflike likelihood (you will need CosmoRec to be installed and camb aware of it, see installation instructions) by doing

cobaya-run /where/to/clone/examples/act_dr6_example.yaml -p /where/to/put/packages

which should run a MCMC sampler for ACT DR6 official data file (i.e. dr6_data.fits in the act_dr6_example.yaml file) using the combination of TT, TE and EE spectra (i.e. polarizations: ['TT', 'TE', 'ET', 'EE']). The results will be stored in the chains/mcmc directory.