GoalRegionSampler.m

% GoalRegionSampler samples the goal region using
% halton sequencing
%
%   Ross Allen, ASL, Stanford University
%   May 19, 2014
%
%   Notes:
%       - Current version assumes goal region is a hyper-rectangle
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

function [ mpinfo ] = GoalRegionSampler( mpinfo )

% Unpack variables to be accessed and modified
nGoalSamples = mpinfo.sampling.nGoalSamples;

% Extract goal region from vicon(assumed second portion of buffer)
% expected form: [start state (nStateDims), goal region (2*nStateDims), active obsacles (4*nActiveObs)]
if (mpinfo.comms.recvGoalRegion)
    mpinfo.termStates.goalRegion = reshape(mpinfo.comms.recvBuffer(...
        mpinfo.systemDefs.nStateDims+1:3*mpinfo.systemDefs.nStateDims),...
        2,mpinfo.systemDefs.nStateDims)';
end

% Unpack variables to be accessed (not modified)
samplingRange = mpinfo.termStates.goalRegion;
sysdefs = mpinfo.systemDefs;
mpinfo.nSamples = mpinfo.sampling.nSamples;

halton_skip = randi(1e6,1); % randomly generate number of initial halton sequence to omit

Xgoal = HaltonSampling(...
    sysdefs.nStateDims, nGoalSamples, samplingRange, halton_skip, 0, true);

% Consolidate results
mpinfo.termStates.Xgoal = Xgoal;

end