glmc.control {glmc}R Documentation

Auxiliary for Controlling GLM Fitting with population level constraints.

Description

Auxiliary function as user interface for glmc fitting. Typically only used when calling glmc.

Usage

glmc.control(epsilon.glm = 1e-8, maxit.glm= 100, trace.glm= FALSE, 
             trace.optim= 0, fnscale.optim=-1, parscale.optim = rep.int(1,1), 
             ndeps.optim = rep.int(0.001,1), maxit.optim = 100, 
             abstol.optim = -Inf, reltol.optim= sqrt(.Machine$double.eps), 
             alpha.optim = 1, beta.optim = 0.5, gamma.optim = 2, 
             REPORT.optim= 10, type.optim = 1, lmm.optim = 5, 
             factr.optim= 1e+07, pgtol.optim = 0, tmax.optim = 10, 
             temp.optim =10, maxit.weights = 25, gradtol.weights = 1e-07, 
             svdtol.weights = 1e-09, itertrace.weights = FALSE)

Arguments

epsilon.glm

positive convergence tolerance epsilon; the iterations converge when |dev - devold|/(|dev| + 0.1) < epsilon.

maxit.glm

integer giving the maximal number of IWLS iterations.

trace.glm

logical indicating if output should be produced for each iteration.

trace.optim

Non-negative integer. If positive, tracing information on the progress of the optimization is produced. Higher values may produce more tracing information: for method “L-BFGS-B” there are six levels of tracing. (To understand exactly what these do see the source code: higher levels give more detail.)

fnscale.optim

A negative number determining the overall scaling to be applied to the value of fn and gr during optimization. In glmc optimization is performed on fn(par)/(fnscale.optim) if fnscale.optim is negative and on fn(par)/((-1)*fnscale.optim) if fnscale.optim is positive.

parscale.optim

A vector of scaling values for the parameters. Optimization is performed on par/parscale and these should be comparable in the sense that a unit change in any element produces about a unit change in the scaled value.

ndeps.optim

A vector of step sizes for the finite-difference approximation to the gradient, on par/parscale scale. Defaults to 1e-3.

maxit.optim

The maximum number of iterations. Defaults to 100 for the derivative-based methods, and 500 for “Nelder-Mead”. For “SANN” maxit gives the total number of function evaluations. There is no other stopping criterion. Defaults to 10000.

abstol.optim

The absolute convergence tolerance. Only useful for non-negative functions, as a tolerance for reaching zero.

reltol.optim

Relative convergence tolerance. The algorithm stops if it is unable to reduce the value by a factor of reltol * (abs(val) + reltol) at a step. Defaults to sqrt(.Machine\$double.eps), typically about 1e-8.

alpha.optim, beta.optim, gamma.optim

Scaling parameters for the “Nelder-Mead” method. alpha is the reflection factor (default 1.0), beta the contraction factor (0.5) and gamma the expansion factor (2.0).

REPORT.optim

The frequency of reports for the “BFGS” and “L-BFGS-B” methods if control\$trace is positive. Defaults to every 10 iterations.

type.optim

for the conjugate-gradients method. Takes value 1 for the Fletcher–Reeves update, 2 for Polak–Ribiere and 3 for Beale–Sorenson.

lmm.optim

is an integer giving the number of BFGS updates retained in the “L-BFGS-B” method, It defaults to 5.

factr.optim

controls the convergence of the “L-BFGS-B” method. Convergence occurs when the reduction in the objective is within this factor of the machine tolerance. Default is 1e7, that is a tolerance of about 1e-8.

pgtol.optim

helps controls the convergence of the “L-BFGS-B” method. It is a tolerance on the projected gradient in the current search direction. This defaults to zero, when the check is suppressed.

temp.optim

controls the “SANN” method. It is the starting temperature for the cooling schedule. Defaults to 10.

tmax.optim

is the number of function evaluations at each temperature for the “SANN” method. Defaults to 10.

maxit.weights

an optional integer to control iteration when solve constrained maximisation for the weights.

gradtol.weights

an optional real value for convergence test while calculating the weights.

svdtol.weights

an optional real value to detect singularity while solve equations. This is used to compute the weights.

itertrace.weights

a logical value. If the iteration history when calculating the weights needs to be printed out.

Value

A list with components


[Package glmc version 0.3-1 Index]