function [v,h]=studmax(beta,PT,PS,R);
%This function solves the student problem for PT as TV transition, PS as
%study transition, and R as the possible returns.  The rows of R correspond
%to the initial states.  The columns of R correspond to the two possible
%actions, TV and study.  The output, v, is the correct value function.  The
%output, h, is a row consisting of 1's when the optimal action is TV and
%2's when the optimal action is study.  Beta is the discount factor.
v0=[0 0 0];
m=1;
while m>0.0001;
A=R+beta*[(PT*v0')';(PS*v0')'];
[v1,index]=max(A);
m=norm(v1-v0,2);
v0=v1;
end
v=v0
h=index