fitCyclicMLA: Fits the cyclic model on ESM data.

In PeterVerboon/cyclicpkg: Analysis of (ESM/EMA) models with cyclic terms

Description

This function fits the cyclic model on ESM data, using a multilevel model (see Verboon & Leontjevas, 2018)

Usage

fitCyclicMLA(dat, yvar = NULL, xvar1 = NULL, xvar2 = NULL,
  id = NULL, cov = NULL, ncycle = 1, P = NULL, P2 = NULL,
  random = "intercept")

Arguments

dat	data frame containing variables
yvar	dependent variable
xvar1	time variabele indicator of first or only cycle
xvar2	time variabele indicator of second (longer) cycle
id	group (subjects, clusters) identifier
cov	additional variabels in the model
ncycle	is the number of cyclic processes in the model (1 = 1, otherwise = 2)
P	is the periodicity of the first cyclic process
P2	is the periodicity of the second cyclic process
random	keyword indicating which terms should be random effect in the model. The following keywords are available: "intercept"= intercept only; "first"= cyclic parameters of first cyclic process; "second" = cyclic parameters of second cyclic process; "cov" = covariates only, no cyclic terms; and finally "all" = all terms

Value

list containing the following elements:

parameters = intercept, amplitude and phase (and other terms in the model)

fit = object from lmer()

formula = specification of lmer model

period = specified periodicity of cyclic process

Examples

 data("smokedat")
  model <- fitCyclicMLA(dat=smokedat,  yvar="intention", xvar1="beepnr",
  xvar2="daynr", id = "subjnr", random = "all",
  ncycle = 1, cov = c("stress", "positiveAffect"))

PeterVerboon/cyclicpkg documentation built on March 4, 2020, 1:27 p.m.