monpol.control {MonoPoly}

Control the Iterations in monpol

Description

Allow the user to set some characteristics of the monpol monotone polynomial fitting algorithm.

Usage

monpol.control(maxiter = 1000, tol = 1e-05,
               tol1=1e-10, tol2=1e-07, tolqr=1e-07)

Arguments

maxiter

A positive integer specifying the maximum number of iterations allowed, used in all algorithms.

tol

A positive numeric value specifying an absolute tolerance for determining whether entries in the gradient are zero for algorithms ‘Full’, ‘BCD’, ‘CD1’ and ‘CD2’.

tol1

A positive numeric value, used in algorithm ‘Hawkins’. Any number not smaller than -tol1 is deemed to be non-negative.

tol2

A positive numeric value, used in algorithm ‘Hawkins’. Any number whose absolute value is smaller than tol2 is taken to be zero.

tolqr

A positive numeric value, used in algorithm ‘Hawkins’ as tolerance for the QR factorisation of the design matrix.

Value

A list with exactly five components:

maxiter

tol

tol1

tol2

tolqr

with meanings as explained under ‘Arguments’.

Author(s)

Berwin A Turlach <[e-mail]>

See Also

monpol, monpol.fit, qr

Examples

monpol.control(maxiter = 2000)
monpol.control(tolqr = 1e-10)