bayes_control: Bayesian closed-loop control

Description Usage Arguments Value

View source: R/simulations.R

Description

Function to provide Bayesian closed-loop control.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
bayes_control(
  targets,
  updates,
  prior,
  true_pars,
  pkmod = pkmod3cptm,
  pdmod = emax_eleveld,
  pdinv = inv_emax_eleveld,
  init0 = NULL,
  init_p = NULL,
  obs_tms = NULL,
  dt_obs = 1/6,
  sim_starttm = 0,
  tci_alg = "effect",
  print_progress = FALSE
)

Arguments

targets

Data frame with columns ("time","target")

updates

Data frame of times at which closed-loop updates should be conducted and optional variable with logical values named 'full_data' indicating if full updates should be used. Defaults to partial.

prior

List with elements "mu" and "sig" specifying the prior mean and covariance matrices for the logged parameter values.

true_pars

Vector of true patient PK-PD parameters.

pkmod

PK model

pdmod

PD model

pdinv

Inverse PD model

init0

True initial concentrations

init_p

Predicted initial concentrations

obs_tms

Times at which observations are collected. If null, observations will be made at fixed intervals specified by 'dtm'.

dt_obs

Interval between measurements.

sim_starttm

Start time of simulation

tci_alg

TCI algorithm used. Defaults to effect-site targeting.

print_progress

Logical. Should current update times be printed to the console.

Value

Returns a list with class "bayessim" containing results from the Bayesian closed-loop simulation. A "plot.bayessim" method exists for visualizing results.


tci documentation built on Feb. 26, 2021, 5:07 p.m.