Main Content

sbiofitstatusplot

Plot status of nonlinear mixed-effects estimation

Syntax

stop = sbiofitstatusplot(beta, status, state)
stop = sbiofitstatusplot(beta, status, state, fenames)

Description

stop = sbiofitstatusplot(beta, status, state) initializes or updates a plot with the fixed effects, beta, the log likelihood status.fval, and the variance of the random effects, diag(status.Psi).

The function returns an output (stop) to satisfy requirements for the 'OutputFcn' option of nlmefit or nlmefitsa. For sbiofitstatusplot, the value of stop is always false.

stop = sbiofitstatusplot(beta, status, state, fenames) specifies the names for the fixed-effects fenames.

Use sbiofitstatusplot to obtain status information about NLME fitting when using the sbiofitmixed function. Specify @sbiofitstatusplot for the 'OutputFcn' field of a statset option structure and then pass in the structure as an input argument to sbiofitmixed.

Alternatively, you can set the 'ProgressPlot' name-value pair argument to true when you run sbiofitmixed. The function sbiofitmixed then calls sbiofitstatusplot at each function iteration. For details, see Progress Plot.

Input Arguments

beta

Current fixed effects

status

Structure containing several fields

FieldValue
inner

Structure describing the current status of the inner iterations within the ALT and LAP procedures, with the fields:

  • procedure

    • 'PNLS', 'LME', or 'none' when the procedure is 'ALT'

    • 'PNLS', 'PLM', or 'none' when the procedure is 'LAP'

  • state'init', 'iter', 'done', or 'none'

  • iteration — Integer starting from 0, or NaN

procedure'ALT' or 'LAP'
iterationInteger starting from 0
fvalCurrent log-likelihood
PsiCurrent random-effects covariance matrix
thetaCurrent parameterization of Psi
mseCurrent error variance

state

Either 'init', 'iter', or 'done'.

fenames

Character vector, string, string vector, or cell array of character vectors specifying the names of fixed effects

Examples

Obtain status information for NLME fitting:

% Create a statset option with 'OutputFcn'.
fitOptions.Options = statset('OutputFcn',@sbiofitstatusplot);
% Pass the structure to sbiofitmixed function.
results = sbiofitmixed(..., fitOptions);

More About

collapse all

Alt

Alternating algorithm for the optimization of the LME or RELME approximations

FO

First-order estimate

FOCE

First-order conditional estimate

LAP

Optimization of the Laplacian approximation for FO or FOCE

LME

Linear mixed-effects estimation

NLME

Nonlinear mixed effects

PLM

Profiled likelihood maximization

PNLS

Penalized nonlinear least squares

RELME

Restricted likelihood for the linear mixed-effects model

Version History

Introduced in R2009b

See Also

| (Statistics and Machine Learning Toolbox) | | |