BAUMGARTNER2014_LIKELISTAT - Likelihood statistics for evaluation of model performance

Usage

[la,le,ci,lr,pvalue] = baumgartner2014_likelistat(p,tang,rang,target,response,varargin)

`p`	pdf matrix
`tang`	polar angles of possible target angles
`rang`	polar angles of possible response angles
`target`	target polar angles of localization test
`response`	response polar angles of localization test
`varargin`	Use 'normalize' for normalization of likelihoods. 1 corresponds to unitary likelihood. This is the default. Use 'original' according to Langendijk et al. (2002).

`la`	actual likelihood
`le`	expected likelihood
`ci`	99% confidence interval for expected likelihood
`lr`	reference likelihoods

baumgartner2014_likelistat calculates the likelihood statistics for the evaluation of the model performance.

The reference likelihoods are given in the following order:

`'1st'dim`	unimodal (1 gaussian dist.: std=17 deg, mu=0)
`'2nd'dim`	bimodal (2 gaussians: mu1=0, mu2=180)
`'3rd'dim`	trimodal (mu1=0, mu2=90, mu3=180)
`'4th'dim`	unitary