stepit

Using cubic interpolation, a new step size is computed for minimizing a 
function value.

These functions were developed to support functional data
analysis as described in Ramsay, J. O. and Silverman, B. W.
(2005) Functional Data Analysis. New York: Springer and in
Ramsay, J. O., Hooker, Giles, and Graves, Spencer (2009).
Functional Data Analysis with R and Matlab (Springer).
The package includes data sets and script files working many examples
including all but one of the 76 figures in this latter book. Matlab versions
are available by ftp from
<https://www.psych.mcgill.ca/misc/fda/downloads/FDAfuns/>.

S by Jim Ramsey

Functional Data Analysis

James Ramsay

Giles Hooker

Spencer Graves

stepit function

<dl> <dt>linemat</dt>
<dd>A 3 by 5 matrix containing step sizes, function values and slopes at
 multiple values.</dd> <dt>ips</dt>
<dd>If 1, previous slope was positive.</dd> <dt>dblwrd</dt>
<dd>Vector of length 2: dblwrd[1] TRUE means step halved
 dblwrd[2] TRUE means step doubled.</dd> <dt>MAXSTEP</dt>
<dd>maximum allowed size of a step.</dd></dl>

Arguments

Compute a new step size for a linear search within an optimization algorithm. — stepit

<dl>

 <dt>linemat</dt>
<dd>A 3 by 5 matrix containing step sizes, function values and slopes at
 multiple values.</dd>

 <dt>ips</dt>
<dd>If 1, previous slope was positive.</dd>

 <dt>dblwrd</dt>
<dd>Vector of length 2: dblwrd[1] TRUE means step halved
 dblwrd[2] TRUE means step doubled.</dd>

 <dt>MAXSTEP</dt>
<dd>maximum allowed size of a step.</dd>

</dl>

stepit: Compute a new step size for a linear search within an optimization algorithm.

Description

Usage

Value

Arguments

References