This repository contains the likelihood module for the KiDS+VIKING-450 (in short: KV450) correlation function measurements from Hildebrandt et al. 2018 (arXiv:1812.06076).
The module will be working 'out-of-the-box' within a MontePython and CLASS (version >= 2.6 and including the HMcode module) setup. The required KiDS+VIKING-450 data files can be downloaded from the KiDS science data webpage and the parameter file for reproducing the fiducial run of Hildebrandt et al. 2018 (arXiv:1812.06076) is supplied in the subfolder
INPUT within this repository.
Assuming that MontePython (with CLASS version >= 2.6 and including the HMcode module) is set up (we recommend to use the MultiNest sampler!), please proceed as follows:
- Clone this repository
git clone https://github.com/fkoehlin/kv450_cf_likelihood_public.git
kv450_cf_likelihood_public.datafrom this repository into a folder named
(you can rename the folder to whatever you like, but you must use this name then consistently for the whole likelihood which implies to rename the
*.data-file, including the prefixes of the parameters defined in there, the name of the likelihood in the
__init__.py-file and also in the
Set the path to the data folder (i.e.
KV450_COSMIC_SHEAR_DATA_RELEASEfrom the tarball available from the KiDS science data webpage in
kv450_cf_likelihood_public.dataand modify parameters as you please (note that everything is set up to reproduce the fiducial run with
Start your runs using e.g. the
kv450_cf.paramsupplied in the subfolder
INPUTwithin this repository.
Contribute your developments/bugfixes to this likelihood (please use a dedicated branch per fix/feature).
If you publish your results based on using this likelihood, please cite Hildebrandt et al. 2018 (arXiv:1812.06076) and all further references for the KiDS+VIKING-450 data release (as listed on the KiDS science data webpage) and also all relevant references for Monte Python and CLASS.
run_with_multinest.sh within the subfolder
INPUT for all MultiNest-related settings that were used for the fiducial runs.
Note when you run the likelihood for the very first time, the covariance matrix from the data release (given in list format) needs to be converted into an actual NxN matrix format. This will take several minutes, but only once. The reformatted matrix will be saved to and loaded for all subsequent runs of the likelihood from the folder
FOR_MONTE_PYTHON within the main folder
KV450_COSMIC_SHEAR_DATA_RELEASE of the data release.
WARNING: This likelihood only produces valid results for
\Omega_k = 0, i.e. flat cosmologies!
For questions/comments please use the issue-tracking system!