sim_ema: Simulate EMA data

In jruwaard/emaph: The accompanying R package of the manual "Ecological Momentary Assessment in Mental Health Research"

Description

Simulate EMA data

Usage

sim_ema(plan = sample_plan(), mm_par = list(fixed = c(intercept = 0,
  time = 0, time2 = 0), random = c(intercept = 0, time = 0, time2 = 0),
  error = 0.02, phi = 0), lim = NULL)

Arguments

plan	a sample plan, as generated by 'sample_plan'
mm_par	a list, defining the parameters of a mixed effects model; the list should contain elements "fixed", "random", "error" and "phi".
lim	optional upper and lower limits of data (defaults to NULL, in which case data values are unbounded).

Value

a data.frame, containing the simulated data:

id	Participant id (numeric)
observation	observation counter (within participant; numeric)
day	day counter (numeric)
beep	beep counter, within day (numeric)
t	time of day (string, in "%H:%M" format)
Y	Simulated EMA value

Examples

sample_plan <- sample_plan(n_participants = 20, n_days = 7)

mm_par_a <- list(
  fixed  = c(intercept = 0,
             time      = 0),
  random = c(intercept = 0.1,
             time      = 0.005),
  error  = 0.2,
  phi    = 0.5
)
a <- sim_ema(plan = sample_plan,
             mm_par = mm_par_a)

mm_par_b <- mm_par_a
mm_par_b$fixed['time']   <- -0.5
mm_par_b$random['time']  <- 0.005

b <- sim_ema(plan = sample_plan,
             mm_par = mm_par_b)

b$id <- b$id + max(a$id)

c <- rbind(a, b)
c$group <- c(rep(0, nrow(a)), rep(1, nrow(b)))

# plot
library(ggplot2)
ggplot(data = c, aes(x = time, y = Y)) +
       geom_line(aes(group = id, colour = factor(group))) +
       geom_point(size = .8) +
       geom_smooth(aes(group = factor(group), colour = factor(group)),
                   method = "loess", se = FALSE, size = 2) +
       xlab("Time")

# test
library(nlme)
fm <- lme(Y ~ 1 + time * group ,
         random = ~ 1 + time | id,
         data = c,
         correlation = corAR1())
summary(fm)

jruwaard/emaph documentation built on May 3, 2019, 8:04 p.m.