simple_like

Generate a simple 2pt likelihood given data, theory and covariance with no cuts

File	likelihood/2pt/cosebis/simple_like.py
Attribution	Chieh-An Lin
	Tilman Troester
	Marika Asgari
	Catherine Heymans
URL

This is a simple likelihood calculation that was set up for COSEBIs. In principle it could be used for any statistic

provided you don’t want to make any cuts to the data, theory and covariance (If you want to do something more complex use 2pt_like.py). It first reads in the tomographic data vector and covariance matrix from the block. The theory is then also extracted and ordered according to the CosmoSIS convention. The likelihood (data-model)^T C^{-1} (data-model) is returned.

Assumptions

• The tomographic data and covariance are ordered in the standard CosmoSIS format

Setup Parameters

Name	Type	Default	Description
data_set	str	En n	Space-separated list of statistic name and angle name to look in for data
data_file	str		Name of FITS file containing the data and covariance matrix
like_name	str	cosebis	The name of the likelihood to save e.g. ‘like_name = cosebis’ will return cosebis_like to the likelihood block
theory_section	str	cosebis	Name of the block section that is storing the theory that you want to compare the data_set to

Input values

Section	Name	Type	Default	Description
cosebis	En	real 1d		The theory predictions
	n	real 1d		If cosebis used, the COSEBIs mode
	bin_{i}_{j}	real 1d		For various i,j depending what is found in the file, the theory predictions for this value.

Output values

Output values

Section	Name	Type	Description
likelihoods	cosebis_like	real	Gaussian likelihood value. Name can be changed in parameter file (see above).