Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!swrinde!elroy.jpl.nasa.gov!jato!kalessin!sam
From: sam@kalessin.jpl.nasa.gov (Sam Sirlin)
Newsgroups: comp.lang.apl
Subject: Re: WANTED: APL maximization function
Message-ID: <1991Mar31.212330.28955@jato.jpl.nasa.gov>
Date: 31 Mar 91 21:23:30 GMT
References: <20114@brahms.udel.edu>
Sender: news@jato.jpl.nasa.gov
Organization: Jet Propulsion Laboratory, Pasadena, CA
Lines: 18
Nntp-Posting-Host: kalessin.jpl.nasa.gov

In article <20114@brahms.udel.edu> fmaster@brahms.udel.edu (Fred A Masterson) writes:
>I am seeking an APL function that will maximize a function of
>one variable, i.e., find the value of x that maximizes f(x).  

To do this once, a simple plot is pretty powerful. To automate this,
you need to decide a bunch of things such as 

  - is the function differentiable (analytically or numerically) to
    some order?
  - is there a single maximum (no local maxima)?
  - do you know a priori bounds on a (finite) region that contains the
    desired maxima?

Given the last condition, a simple binary search works pretty fast and
is simple to code (I have one somewhere).

Sam Sirlin
Jet Propulsion Laboratory         sam@kalessin.jpl.nasa.gov