Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: RE: code for dynamic programming

From   Nick Cox <[email protected]>
To   "'[email protected]'" <[email protected]>
Subject   RE: st: RE: code for dynamic programming
Date   Thu, 1 Mar 2012 15:28:40 +0000

I don't know what your expectations are here. 

If someone has already done this in Stata, or something very similar, then they will have code that they can post, and your luck is in. 

Otherwise, the prescription "optimization with iteration" is a pretty large fraction of what Stata does. The most general tools are part of Mata. 

[email protected] 

Osei Kuffour Wiafe

Thanks, I have seen the FORTRAN, looks fantastic but not much help for
my current need.Maybe a little explanation of the nature of the
equation might help.

It's a dynamic optimization problem which can be solved by backward
recursion, where I find the expectation at the time of a future
terminal date (which is known) and iterate backwards to the present.

Any optimization with iteration should help me out! Any ideas coming up?

On Thu, Mar 1, 2012 at 9:22 PM, Nick Cox <[email protected]> wrote:

> The code for -group1d- (SSC) (that's "1" (one) not the letter "l") uses a
> dynamic programming algorithm courtesy of John Hartigan's original FORTRAN,
> translated into my FORTRAN translated into some BASIC translated into
> Stata/Mata.
> I don't know what the Epstein-Zin utility is and I give no promises or
> predictions about how much it may help.
> If you're quite new to Stata this sounds a bit tough for a first
> programming problem. If you have expertise in one or more programming
> languages, much less so.

Osei Kuffour Wiafe

> I'm quite new to stata and I need help with coding a backward recursive
> problem. It is in the form of the Epstein-Zin utility and similar to the
> Bellman equation. I'd appreciate any other dynamic programming code which
> would relate to what I am doing. That should give me a clue to finding my
> solution. For now I have no idea where to begin!

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index