GetRewardModelFixedA

PURPOSE

Defines the reward function for a given action.

SYNOPSIS

function ra=GetRewardModelFixedA(RM,a)

DESCRIPTION

   Defines the reward function for a given action.

   Particularizes a reward function defined on continuous state and action
   spaces for a particular action 'a'.

   The particularization is obtained by evaluating the Gaussians in 'a'
   given when defining the model. The result is a Gaussian mixture in 's'.

   See also CS_CA_RewardModel.

CROSS-REFERENCE INFORMATION

This function calls:

GMixture Gaussian mixture constructor.
Value Evaluates a GMixture.
Value Evaluation of a Gaussian.

This function is called by:

MakeFigure1 Generates Figure 1.
ComputeAlpha_a Compute the alpha_i_n-element for the given action and belief.
ComputeAlpha_a Compute the alpha_i_n-element for the given action and belief.
ComputeAlpha_a Compute the alpha_i_n-element for the given action and belief.

SOURCE CODE

0001 function ra=GetRewardModelFixedA(RM,a)
0002 %   Defines the reward function for a given action.
0003 %
0004 %   Particularizes a reward function defined on continuous state and action
0005 %   spaces for a particular action 'a'.
0006 %
0007 %   The particularization is obtained by evaluating the Gaussians in 'a'
0008 %   given when defining the model. The result is a Gaussian mixture in 's'.
0009 %
0010 %   See also CS_CA_RewardModel.
0011 
0012   w=RM.w.*cellfun(@(x)(Value(x,a)),RM.gA);
0013   ra=GMixture(w,RM.gS);
0014   
0015