Institut de Robòtica i Informàtica Industrial

MakeFigureSimulation

PURPOSE ^

Generates the plots in Figures 2, 8, 10, and 11.

SYNOPSIS ^

function MakeFigureSimulation(title,data)

DESCRIPTION ^

   Generates the plots in Figures 2, 8, 10, and 11.

   A function to generate the Figures 2, 8, 10, and 11, showing four
   plots with the evolution in time of

      - The average value per belief (in the paper we actually show the
        sum of values for all beliefs).
      - The average accumulated discounted reward.
      - The number of alpha-elements in the policy.
      - The average number of policy changes per belief (in the paper we
        actually show the total number of policy changes).

   This function does not compute anything, it just load the file with the
   average results and make the plots.
   The results are loaded for a file with name 
      data-resutls.mat
   where data is the given parameter.

   See also GetPOMDPSolutionStatistics.

CROSS-REFERENCE INFORMATION ^

This function calls:
  • set Set method for Gaussian mixtures.
  • set Set function for PODMP (base type)
This function is called by:

SOURCE CODE ^

0001 function MakeFigureSimulation(title,data)
0002 %   Generates the plots in Figures 2, 8, 10, and 11.
0003 %
0004 %   A function to generate the Figures 2, 8, 10, and 11, showing four
0005 %   plots with the evolution in time of
0006 %
0007 %      - The average value per belief (in the paper we actually show the
0008 %        sum of values for all beliefs).
0009 %      - The average accumulated discounted reward.
0010 %      - The number of alpha-elements in the policy.
0011 %      - The average number of policy changes per belief (in the paper we
0012 %        actually show the total number of policy changes).
0013 %
0014 %   This function does not compute anything, it just load the file with the
0015 %   average results and make the plots.
0016 %   The results are loaded for a file with name
0017 %      data-resutls.mat
0018 %   where data is the given parameter.
0019 %
0020 %   See also GetPOMDPSolutionStatistics.
0021 %
0022 
0023   h=clf;
0024   set(h,'name',title,'numbertitle','off');
0025 
0026   if ischar(data)
0027     Results=load([data '-results.mat']);
0028   else
0029     Results=data;
0030   end
0031   
0032   subplot(2,2,1);
0033   errorbar(Results.tics,Results.SM.AV,Results.SD.AV);
0034   xlabel('time (s)');
0035   ylabel('Value')
0036   
0037   subplot(2,2,2);
0038   errorbar(Results.tics,Results.SM.R,Results.SD.R);
0039   xlabel('time (s)');
0040   ylabel('Acc. Disc. Reward')
0041   
0042   subplot(2,2,3);
0043   errorbar(Results.tics,Results.SM.nAlpha,Results.SD.nAlpha);
0044   xlabel('time (s)');
0045   ylabel('No. \alpha-Elements');
0046   
0047   
0048   subplot(2,2,4);
0049   errorbar(Results.tics,Results.SM.nChanges,Results.SD.nChanges);
0050   xlabel('time (s)');
0051   ylabel('No. Policy Changes');
0052


Institut de Robòtica i Informàtica Industrial

Generated on Wed 05-Aug-2009 15:05:21 by m2html © 2003