simData: Simulate data to fit methods in the R package 'psbpHMM'

In lvhoskovec/psbpHMM: Covariate-dependent iHMMs for multiple time series

Description

Simulate multiple time series data from a fixed-state MVN hidden Markov model

Usage

simData(n = 20, t.max = 288, K = 6, p = 3, trend = "shared", missingLevel = 0)

Arguments

n	number of time series
t.max	length of each time series
K	number of clusters to be shared across all time series
p	number of exposures
trend	"shared" (default) for shared temporal trends or "distinct" for distinct temporal trends among n time series
missingLevel	missing data level, data will be removed and split between missing at random and below LOD

Value

a list with components

• y: scaled simulated data with missing observations set to NA for MAR or -Inf for below LOD

• y.complete: scaled complete data

• z.true: true hidden state trajectories for each time series

• K: true number of hiddens states

• mu.true: scaled true state-specific means

• lod: limit of detection

lvhoskovec/psbpHMM documentation built on Feb. 13, 2022, 10:40 p.m.