Calibrate prior distributions based on quartiles

Source: R/priors_calibration.R

Calibrates prior distributions for parametric survival analysis based on median survival and interquartile range. Returns a list of prior distribution for the intercepts and auxiliary parameters.

calibrate_quartiles(
  median_t,
  iq_range_t,
  prior_sd = 0.5,
  distributions = c("exp-aft", "weibull-aft", "lnorm-aft", "llogis-aft", "gamma-aft"),
  verbose = FALSE,
  search_bounds1 = c(-100, 100),
  search_bounds2 = c(0 + 0.01, 100)
)

Arguments

median_t

median survival

iq_range_t

interquartile range of the survival

prior_sd

pre-specified standard deviation of the prior distributions (either a single value that is used for both the intercept and auxiliary parameter or a vector where the first value corresponds to the sd for the prior distribution on the intercept and the second value to the sd for the prior distribution on the auxiliary parameter)

distributions

vector of parametric families for which prior distributions ought to be calibrated

verbose

whether debug information be printed

search_bounds1

search boundaries for the intercept parameter

search_bounds2

search boundaries for the auxiliary parameter

Value

returns a list of prior distribution for the intercepts and auxiliary parameters.

Examples

priors <- calibrate_quartiles(median_t = 5, iq_range_t = 10, prior_sd = 0.5)