function [s]=convert_values_to_state(x,y,vx,vy)global VMAX Map_Data Pos_Vector_Indexes Speed_Vector_Indexes% convert_values_to_state: Converts a [x y vx vy] representation to the%                              singleton [s] state representation.% Arguments --------------------------------------------------------------%     [x,y] = valid coordinates of the car with respect to X and Y axis %     [vx,vy] = valid speed of the car with respect to X and Y axis % Evaluation -------------------------------------------------------------%     s = the converted state element; as there is a bijection between both%         representations, it is unique for a given [x,y,vx,vy]%--------------------------------------------------------------------------% In verbose mode, there is no difference to silent mode.%--------------------------------------------------------------------------% MDP Toolbox, INRA, BIA Toulouse, France%--------------------------------------------------------------------------% Checking dimensions of the vectorsif size(x,2)==1    x=x';endif size(y,2)==1    y=y';end% We will first determine the speed_index and position_index, that is the % place occupied by the position and speed of the car within the ordered% set of valid possibilities. Due to the structure of the state vector,% we have position_index and speed_index equal to the indexes i and j where % Pos_Vector_Indexes(i)=(x-1)*m+y% Speed_Vector_Indexes(j)=(vx-1)*(2*VMAX+1)+vy% Caution: vx, vy are in the -VMAX:VMAX range, so we will need to recenter% them to 1:2*VMAX+1m=size(Map_Data,1);position_index=zeros(size(x));speed_index=zeros(size(x));for t=1:size(x,2),    position_index(t)=find(Pos_Vector_Indexes==((x(t)-1)*m+y(t)));    speed_index(t)=find(Speed_Vector_Indexes==((vx(t)+VMAX)*(2*VMAX+1)+vy(t)+VMAX+1));end% We can now compute s, as s=n_speed*(position_index-1)+speed_index s=zeros(size(x));exist=(x>0 & x<=size(Map_Data,2) & y>0 & y<=size(Map_Data,1));s(exist)=size(Speed_Vector_Indexes,1)*(position_index(exist)-1)+speed_index(exist);s(~exist)=0;