Constructs the covariance matrix for multiple measurements of the same cluster if the same individuals are observed at all time periods. This function is not designed to be used directly.
construct_CovSubMat(
N,
timepoints,
sigma,
tau,
eta = NULL,
tauAR = NULL,
etaAR = NULL,
rho = NULL,
gamma = 0,
trtMat = NULL,
psi = NULL,
INDIV_LVL = FALSE
)Number of individuals per cluster
numeric (scalar or vector), number of timepoints (periods). If design is swd, timepoints defaults to length(Cl)+1. Defaults to 1 for parallel designs.
numeric (vector of length `timepoints`), residual error
numeric (vector of length `timepoints`), standard deviation of random intercepts
numeric (vector of length `timepoints`), standard deviation of random slope
numeric (scalar), value between 0 and 1. Defaults to NULL. If `tauAR` is not NULL, the random intercept `tau` is AR1-correlated. *Currently not compatible with `rho`!=0 !*
numeric (scalar), value between 0 and 1. Defaults to NULL. If `etaAR` is not NULL, the random slope `eta` is AR1-correlated. *Currently not compatible with `rho`!=0 !*
numeric (scalar), correlation of `tau` and `eta`
numeric (vector of length `timepoints`), standard deviation of a random time effect.
a matrix of dimension *#Cluster* x *timepoints* as produced by the function `construct_trtMat`, indicating the cluster-periods that receive interventional treatment. Defaults to NULL. If trtMat is given, the arguments `SumCl` and `timepoints` are ignored (!).
numeric (scalar), random subject specific intercept. Leads to a closed cohort setting
logical, should the computation be conducted on an individual level? This leads to longer run time and is mainly for diagnostic purposes.
a block of a covariance matrix with two levels of clustering, corresponding to intra-cluster covariance over time for one cluster