void SetActionValue(const TState& s, const TAction& a, float v) { assert(s.GetID() < mActionValues.size()); assert(a.GetID() < mActionValues[s.GetID()].size()); mActionValues[s.GetID()][a.GetID()] = v; }
float GetActionValue(const TState& s, const TAction& a) const { assert(s.GetID() < mActionValues.size()); assert(a.GetID() < mActionValues[s.GetID()].size()); return mActionValues[s.GetID()][a.GetID()]; }