function m = baumgartner2014_virtualexp(p,tang,rang,varargin)
%BAUMGARTNER2014_VIRTUALEXP Performs a virtual sound-localization experiment
%   Usage:    m = baumgartner2014_virtualexp(p,tang,rang)
%
%   Input parameters:
%     p       : prediction matrix containing probability mass vectors (PMVs) 
%               for the polar response angle as a function of the polar  
%               target angle (1st dim: response angle, 2nd dim: target
%               angle)
%     rang    : polar response angles
%     tang    : polar target angles
%
%   Output parameters:
%     m       : item list of virtual experiment,
%               in the format [azi_target, ele_target, azi_response, ele_response, lat_target, pol_target, lat_response, pol_response, F/B-C resolved pol_response]
%
%
%   BAUMGARTNER2014_VIRTUALEXP(...) runs virtual localization experiments where the
%   response behavior is based on (predicted) polar response PMVs.
%
%
%
%   BAUMGARTNER2014_VIRTUALEXP accepts the following optional parameters:
%
%     'runs',runs    	Define the number of runs. 
%                    	Default value is 10.
%
%     'targetset',ts  Define the set of polar target angles.
%                     As default 'tang' is used.
%
%     'lat',lat     	Define the lateral target angles. 
%                    	Default value is 0 deg.
%
%   See also: localizationerror, baumgartner2014, plot_baumgartner2014
%
%   Url: http://amtoolbox.org/amt-1.4.0/doc/modelstages/baumgartner2014_virtualexp.php


%   #StatusDoc: Perfect
%   #StatusCode: Perfect
%   #Verification: Verified
%   #Requirements: SOFA CircStat M-SIGNAL M-Stats O-Statistics
%   #Author: Robert Baumgartner (2014), Acoustics Research Institute, Vienna, Austria

% This file is licensed unter the GNU General Public License (GPL) either 
% version 3 of the license, or any later version as published by the Free Software 
% Foundation. Details of the GPLv3 can be found in the AMT directory "licences" and 
% at <https://www.gnu.org/licenses/gpl-3.0.html>. 
% You can redistribute this file and/or modify it under the terms of the GPLv3. 
% This file is distributed without any warranty; without even the implied warranty 
% of merchantability or fitness for a particular purpose. 

%   References: baumgartner2014modeling


definput.keyvals.runs = 10;
definput.keyvals.targetset = [];
definput.keyvals.lat = 0;
% definput.flags.colorbar = {'colorbar','no_colorbar'};
[flags,kv]=ltfatarghelper({'runs','targetset'},definput,varargin);

if isempty(kv.targetset)
  kv.targetset = tang;
end


%% Run experiments
nt=length(kv.targetset);
m = nan(nt*kv.runs,9);
m(:,5) = kv.lat;
m(:,6) = repmat(kv.targetset(:),kv.runs,1);
m(:,7) = kv.lat;
% kv.targetset = round(kv.targetset);
if length(tang) > 1
  tangbound = tang(:)+0.5*diff([tang(1)-diff(tang(1:2));tang(:)]);
else
  tangbound = tang;
end
post=zeros(nt,1); % indices of target positions
for ii = 1:nt
  if kv.targetset(ii) > max(tangbound)
    post(ii) = length(tangbound); % if outside predicted range, set to most extreme possible position
  else
    post(ii) = find(tangbound>=kv.targetset(ii),1);
  end
end

posr=zeros(nt,1);
for rr=1:kv.runs
  for jj = 1:nt 
    posr(jj) = local_discreteinvrnd(p(:,post(jj)),1);
    m(jj+(rr-1)*nt,8) = rang(posr(jj));
  end
end



end


function [ X ] = local_discreteinvrnd(p,n,m)
% DISCRETEINVRND implements an inversion method for a discrete distribution
% with probability mass vector p for n trials
% Usage:    [ X ] = discreteinvrnd(p,n)
%
% AUTHOR : Robert Baumgartner

if ~exist('m','var')
    m=1;
end

p = p/sum(p);   % ensure probability mass vector
c = cumsum(p);
t = max(c)*rand(n,m); % rand returns ]0,1]
X = zeros(n,m);
for jj = 1:m
    for ii = 1:n
        X(ii,jj) = find(c >= t(ii,jj) ,1);
    end
end

end