public void Set(int row, int col, double value)
{
var ix = (Columns * row) + col;
Tembo.Assert(ix >= 0 && ix < W.Length);
W[ix] = value;
}
/// <summary>
/// Returns an action from a state
/// </summary>
/// <param name="state">state size must be equal to NumberOfStates</param>
/// <returns></returns>
public int Act(double[] state)
{
Tembo.Assert(state.Length == NumberOfStates, $"Current state({state.Length}) not equal to NS({NumberOfStates})");
var a = 0;
// convert to a Mat column vector
var s = new Matrix(NumberOfStates, 1);
s.Set(state);
// epsilon greedy policy
if (Tembo.Random() < Options.Epsilon)
{
a = Tembo.RandomInt(0, NumberOfActions);
}
else
{
// greedy wrt Q function
var amat = ForwardQ(Network, s, false);
a = Tembo.Maxi(amat.W); // returns index of argmax action
}
// shift state memory
this.s0 = this.s1;
this.a0 = this.a1;
this.s1 = s;
this.a1 = a;
return(a);
}
public double Get(int row, int col)
{
var ix = (Columns * row) + col;
Tembo.Assert(ix >= 0 && ix < W.Length);
return(W[ix]);
}
