baumgartner2014_pmv2ppp - - Performance predictions from PMVs of baumgartner2014

Usage

[ qe,pe,eb ] = baumgartner2014_pmv2ppp( p,tang,rang );
[ qe,pe,eb ] = baumgartner2014_pmv2ppp( p,tang,rang,exptang );
[ qe,pe,eb ] = baumgartner2014_pmv2ppp( pred );

Input parameters

`p`	prediction matrix (response PMVs)
`tang`	possible polar target angles. As default, ARI's MSP polar angles in the median SP is used.
`rang`	polar angles of possible response angles. As default regular 5 deg.-sampling is used (-90:5:265).
`pred`	structure including all input parameters above.

Output parameters

`qe`	quadrant error rate
`pe`	local polar RMS error in degrees
`eb`	elevation bias in degrees; QEs and up-rear quadrant excluded

baumgartner2014_pmv2ppp(...) retrieves commonly used PPPs (Psychoacoustic performance parameters) for sagittal-plane (SP) localization like quadrant error (QE), local polar RMS error (PE), and elevation bias (EB) from response PMVs (probability mass vectors) predicted by a localization model. PPPs are retreived either for a specific polar target angle or as an average across all available target angles. The latter is the default.

baumgartner2014_pmv2ppp needs the following optional parameter in order to retrieve the PPPs for a specific (set of) target angles:

'exptang= exptang experimental polar target angles'

baumgartner2014_pmv2ppp accepts the following flag:

`'print'`	Display the outcomes.
`'QE_PE_EB'`	Compute QE, PE, and EB. This is the default.
`'QE'`	Compute QE.
`'PE'`	Compute PE.
`'EB'`	Compute EB.
`'absPE'`	Compute absolute polar error.
`'chance'`	Compute chance performance for QE and PE.

Example:

To evaluate chance performance of QE and PE use

[qe,pe] = baumgartner2014_pmv2ppp('chance');

References: