PSM (version 0.8-6)

PSM.simulate: Create simulation data for multiple individuals

Description

Simulates data for multiple individuals in a mixed effects model based on stochastic differential equations using an euler scheme.

Usage

PSM.simulate(Model, Data, THETA, deltaTime, longX=TRUE)

Arguments

Model

A list containing the model components either Linear or Non-Linear Model list.*

Data

List with elements described below. No Data$Y is needed as it is generated through the simulation. The number of individuals simulated is equal to length(Data). [object Object],[object Object],[object Object]

THETA

Vector of population parameters

deltaTime

Time Step in the Euler scheme

longX

Boolean. Toggles output of the entire simulated outcome of the states

Value

• The simulated outcome of the model is returned in a list, where each element is the data for an individual.
• X
  Simulated states sampled at time points for measurements
• Y
  Simulated measurements
• Time
  Time points for measurements
• U
  Input vector used in the simulation
• eta
  The random effects used in the simulation
• Dose
  The dose list used in the simulation
• longX
  Entire outcome of simulated states
• longTime
  Time points for longX.

Details

The $\eta_i$ is drawn from the multivariate normal distribution $N(0,\Omega)$. The simulation is an euler based method but for every time interval dt the model is predicted and the states affected by system noise ($\sigma$). The measurements are added an normal error term belonging to $N(0,S)$. The function mvrnorm from the MASS pacakge is used to to generate random numbers fra multivariate normal distributions.

References

Please visit http://www.imm.dtu.dk/psm or refer to the help page for PSM.

See Also

PSM, PSM.estimate, PSM.smooth, PSM.plot, PSM.template

Examples

cat("Examples are included in the package vignette.
")