home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!zaphod.mps.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!dog.ee.lbl.gov!network.ucsd.edu!mvb.saic.com!eafbvax!eafbtems!elston
- Newsgroups: comp.lang.c++
- Subject: Re: SUMMARY: Automatic differentiation
- Message-ID: <1992Jul29.072711.300%boot.decnet@edwards-tems.af.mil>
- From: elston%boot.decnet@edwards-tems.af.mil (Mark Elston, NSI Inc., Edwards AFB CA)
- Date: 29 Jul 92 07:27:10 -0700
- References: <l71d15INNakq@almaak.usc.edu> <1992Jul26.002544.12218@rata.vuw.ac.nz> <l75d3fINNmjl@almaak.usc.edu>
- Organization: T&E Mission Simulator - Edwards AFB, CA
- Lines: 41
-
- In article <l75d3fINNmjl@almaak.usc.edu>, ajayshah@almaak.usc.edu (Ajay Shah) writes:
- > robertd@kauri.vuw.ac.nz (Dr R Davies*xrbd*) writes:
- >
- >>In article <l71d15INNakq@almaak.usc.edu> ajayshah@almaak.usc.edu (Ajay Shah)
- >> writes:
- >
- >>To do symbolic manipulation in C++ you will probably want to use the kind of
- >>representation I use in my matrix package, Newmat. Of course the basic
- >>concept in that representation is taken from the reading I have done about
- >>symbolic manipulation.
- >
- > On to automatic differentiation. No, automatic differentiation is NOT
- > symbolic differentiation. The key point in AD is we never evaluate
- > the functional form of the derivatives. We merely use the rules of
- > calculus to numerically evaluate the derivatives at a given point.
- >
- > ...
- >
- > The broad idea is like this. Suppose we are evaluating a expression
- > f(x,y,z) = x + (y*z). If we overload '*' and '+' then it's possible
- > to maintain the bookkeeping.
- >
- > [the rest deleted]
-
- There is an interesting book that you might want to get hold of. It is called
- "Numerical Derivatives and Nonlinear Analysis" by Harriet Kagiwada, Robert
- Kalaba, Numa Rasakhoo and Karl Spingarn. (Plenum Press, New York, 1986).
-
- The first three chapters of the book deal with implementing this very type of
- solution in FORTRAN and using it in optimal control/calculus of variations
- kinds of problems. They call their implementation FEED (Fast and Efficient
- Evaluation of Derivatives).
-
- The approach that Ajay recommends should work and fits in well with FEED.
-
- ------------------------------------------
- Mark Elston
- elston@edwards-tems.af.mil
-
- "Just trying to shed a little heat on the subject"
-
-