spm_discrete: Discrete multi-dimensional optimization

In stpm: Stochastic Process Model for Analysis of Longitudinal and Time-to-Event Outcomes

Discrete multi-dimensional optimization

Description

Discrete multi-dimensional optimization

Usage

spm_discrete(
  dat,
  theta_range = seq(0.02, 0.2, by = 0.001),
  tol = NULL,
  verbose = FALSE
)

Arguments

dat	A data table.
theta_range	A range of theta parameter (axe displacement of Gompertz function), default: from 0.001 to 0.09 with step of 0.001.
tol	A tolerance threshold for matrix inversion (NULL by default).
verbose	An indicator of verbosing output.

Details

This function is way more faster that continuous spm_continuous_MD(...) (but less precise) and used mainly in estimation a starting point for the spm_continuous_MD(...).

Value

A list of two elements ("dmodel", "cmodel"): (1) estimated parameters u, R, b, Sigma, Q, mu0, theta for discrete-time model and (2) estimated parameters a, f1, Q, f, b, mu0, theta for continuous-time model. Note: b and mu0 from first list are different from b and mu0 from the second list.

References

Akushevich I., Kulminski A. and Manton K. (2005), Life tables with covariates: Dynamic model for Nonlinear Analysis of Longitudinal Data. Mathematical Population Studies, 12(2), pp.: 51-80. <DOI:10.1080/08898480590932296>.

Examples

library(stpm)
data <- simdata_discr(N=10)
#Parameters estimation
pars <- spm_discrete(data)
pars

stpm documentation built on Sept. 5, 2022, 5:06 p.m.