After #1 is complete, you will see that the method uses an iterative step to update the value function. This works, but in principle the derivative information could provide better convergence of the model (and enable writing it in a way that can be extended).+
To do this, we first need to rewrite this as a system of equations in the vector V where a solution is given if the stacked equations are all close to 0. Unlike the existing code, which solves a linear system, this will be written as a nonlinear system of equations).
After #1 is complete, you will see that the method uses an iterative step to update the value function. This works, but in principle the derivative information could provide better convergence of the model (and enable writing it in a way that can be extended).+
To do this, we first need to rewrite this as a system of equations in the vector V where a solution is given if the stacked equations are all close to 0. Unlike the existing code, which solves a linear system, this will be written as a nonlinear system of equations).